DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This Final Office Action is responsive to Applicant's amendment filed on 26 May 2026. Applicant’s amendment on 26 May 2026 amended Claims 1, 8, and 15. Currently Claims 1-3, 5-10, 12-17, and 19-23 are pending and have been examined. Claims 4, 11, and 18 were previously canceled. The Examiner notes that the 101 rejections have been maintained.
Response to Arguments
Applicant's arguments filed 26 May 2026 have been fully considered but they are not persuasive.
The Applicant argues on page 12-13 that the “Office Action alleges that the claims represent a combination of abstract mathematical concepts abstract concepts of organizing human activity, and mental processes Office Action, pages 19-20. However, that improperly parses the claims into multiple exceptions”.
The Examiner respectfully disagrees.
With respect to the argument the Examiner notes that while the Applicant's argument that the Office Action improperly parses the claims into multiple exceptions is not persuasive. The Examiner respectfully maintains the rejection. Specifically, MPEP 2106.04, subsection II.B expressly provides that when a claim recites multiple judicial exceptions whether falling within the same or different groupings examiners should consider those limitations together as a single abstract idea, rather than as a plurality of separate abstract ideas to be analyzed individually. The Office Action did precisely that. The rejection identified claim limitations that fall within the mathematical concepts grouping (e.g., determining load amounts and load scores), the certain methods of organizing human activity grouping (e.g., managing how human agents handle customer service interactions based on workload and performance), and the mental processes grouping (e.g., tracking changes in agent behavior and analyzing effects on agent performance), and then expressly stated that "these limitations are considered together as a single abstract idea." See Office Action. This is the correct procedure under MPEP 2106.04, subsection II.B, and is consistent with the USPTO's AI Subject Matter Eligibility Update Examples, which similarly identify limitations spanning multiple groupings and consolidate them into a single abstract idea for further analysis. Furthermore, MPEP 2106.04(a)(2) itself confirms that the groupings are not mutually exclusive, noting that "some claims recite limitations that fall within more than one grouping or sub-grouping," and instructs examiners to "identify at least one abstract idea grouping, but preferably identify all groupings to the extent possible." Identifying limitations that fall within multiple groupings is therefore not improper parsing it is the proper, thorough application of the USPTO's examination framework. The claim limitations reciting mathematical calculations, management of human work distribution, and observation and evaluation of agent behavior were correctly identified and then treated collectively as a single abstract idea for the purpose of Step 2A Prong Two analysis. The rejection is therefore maintained.
The Applicant argues on pages 14-16 that the as previously noted, “Claim 1 recites an artificial intelligence/machine learning (AI/ML) network trained to handle at least one type of interaction on behalf of an agent or an agent set. Claim 1 is directed to improving the AI/ML network in two respects. First, "an amount of interaction load for the AI/ML network and an amount of attention needed by the agent for the at least one type of interaction" are changed according to "one or more changes in the one or more properties or the behavior of the agent" and the "quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set."
Second, the AI/ML network is improved by "expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time as the agent set and the AI/ML network handle interactions." Expanding the knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions involves self-assessment by the AI/ML network of the "quality of performance of the AI/ML network," which may result in "the AI/ML network ... determining to leave [a] new interaction for the agent to handle based on the complexity" of the new interaction, as determined by the AI/ML network.
As taught in the Applicant's specification, "each [human remote customer service] agent [[210a]-[0210n]] may have different capacities at which the agent can perform different tasks at a good performance level," which "becomes more apparent when multi-tasking and remote interactions are added into the agent's workload." In the claimed embodiments, an AI/ML layer 705 performs a percentage of interactions allocated to one of agents [210a]- [0210n].
According to Claim 1, "one or more properties of one or more interactions associated with an agent" are obtained, "an amount of load for the agent generated by ... interactions associated with the agent is determined based on... properties of the... interactions," and "changes in the... properties or a behavior of the agent" are tracked, up to a time when "a new interaction" is identified. These steps are performed by software including the AIML network executing on one or more processing devices. However, "software is not automatically an abstract idea, even if performance of a software task involves an underlying mathematical calculation or relationship." MPEP 2106.04(a).
Based on an outcome of those software-executed steps, the AI/ML network "trained to handle at least one type of interaction on behalf of the agent or the agent set" is improved by "iteratively performing ... a system optimization operation." The system optimization operation includes "changing... an amount of interaction load for the AI/ML network and an amount of attention needed by the agent" according to the "changes in the one or more properties or the behavior of the agent" and a "quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set," where "load scores for the agent..., each... associated with a measure of computing power for the AI/ML network and agent attention needed for the associated interaction" are determined "based on the one or more properties of the one or more interactions." The system optimization operation also includes "expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time" including the tracked quality of performance. "Based on the complexity of the new interaction," "expanding the knowledge base and the capabilities of the AI/ML network" may result in "leaving the new interaction for the agent to handle."
The Examiner respectfully disagrees.
With respect to the argument the Examiner notes that while the Applicant's argument that Claim 1 is directed to improving the AI/ML network in two respects specifically by dynamically changing the amount of interaction load allocated to the network based on tracked performance quality, and by expanding the network's knowledge base and capabilities to handle additional interaction types is not persuasive. The Examiner respectfully maintains the rejection.
While the Examiner acknowledges and agrees with the general legal principle cited by Applicant that software is not automatically an abstract idea even if it involves an underlying mathematical calculation or relationship (MPEP 2106.04(a)), that principle does not resolve the eligibility question here. The relevant inquiry at Step 2A Prong Two is whether the claim as a whole integrates the recited judicial exception into a practical application by reflecting a genuine improvement to computer technology or another technical field. Specifically, the specification must describe an improvement in technology sufficiently such that it would be apparent to one of ordinary skill in the art, and critically, the claim itself must include the components or steps that provide that improvement. See MPEP 2106.04(d)(1) and 2106.05(a); Intellectual Ventures I LLC v. Symantec Corp.; Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB Sept. 26, 2025) (precedential). The present claims fail this two-part test at the claim level.
Regarding the first asserted improvement dynamically changing the amount of interaction load for the AI/ML network based on tracked quality of performance the claim recites this only as a functional outcome: "changing...an amount of interaction load for the AI/ML network and an amount of attention needed by the agent...based on the tracked one or more changes...and the tracked quality of performance." The claim provides no technical detail regarding the mechanism, algorithm, data structure, or parameter adjustment technique by which this reallocation is accomplished. This is precisely the type of functional, result-oriented claiming that the August 4, 2025 Memorandum identifies as reciting merely "the idea of a solution or outcome" rather than "a particular solution to a problem or a particular way to achieve a desired outcome." The specification likewise describes the ML/AI layer only generically as capable of changing "the amount of attention needed by the agent 210a-210n versus work performed by the ML/AI layer 705 for a specific interaction 215" (Spec. [0075]), without disclosing any specific technical mechanism for how this reallocation occurs. A bare functional assertion of capability, without technical implementation details in either the specification or the claim, is insufficient to establish an improvement to technology. See Affinity Labs of Tex. v. DirecTV, LLC.
Regarding the second asserted improvement expanding the knowledge base and capabilities of the AI/ML network to handle additional interaction types Applicant argues that this constitutes a technological improvement to the AI/ML network itself, including through self-assessment of quality of performance. However, as previously addressed, what Applicant characterizes as a technological improvement to the AI/ML network is in fact an improvement to the abstract idea of automating more types of customer service interactions over time that is, an improvement in the performance of the abstract business function, not an improvement to machine learning technology itself. The specification confirms this, describing the expansion of the AI/ML network's capabilities solely in terms of handling "more types of interactions" and enabling agents to focus on "more advanced or complex interactions" (Spec. [0075]-[0076]). There is no disclosure of any novel training methodology, network architecture modification, parameter adjustment scheme, or knowledge representation structure. As the Federal Register guidance accompanying the 2024 AI-SME Update makes clear, "an improvement in the judicial exception itself is not an improvement in the technology," citing In re Board of Trustees of Leland Stanford Junior University. An AI/ML network that becomes better at performing an abstract task distributing and handling customer service interactions represents improvement in the application of the abstract idea, not improvement to the technology of machine learning itself.
Furthermore, this case is distinguishable from Ex Parte Desjardins, upon which Applicant implicitly relies. In Desjardins, the claims reflected specific improvements to how a machine learning model itself operates namely, training a model to learn new tasks while protecting knowledge about previous tasks to overcome the defined technical problem of "catastrophic forgetting" in continual learning systems and the claims included data structure elements reciting specific adjustments in values to a plurality of performance parameters while preserving prior values. The PTAB found these improvements were "tantamount to how the machine learning model itself would function in operation." Here, by contrast, Claim 1 recites no specific technical problem in machine learning technology (such as catastrophic forgetting), no specific data structures, no specific parameter adjustment mechanisms, and no specific training methodology. The claim recites only that the AI/ML network "expand[s] a knowledge base and capabilities...using data collected over time" a generic functional description that covers any conceivable method of doing so and therefore cannot reflect the specific, claim-level technical improvement required by Desjardins and MPEP 2106.05(a). The rejection is therefore maintained.
The Applicant argues on pages 16-17 that “Claim 1 is directed to improvements to the AI/ML network itself. The Office Action asserts that "under the 2024 AI-SME Guidance, an improvement in the judicial exception itself (improving task distribution methods) is not an improvement in technology." However, the recited improvement is also to a knowledge base and capabilities of the AI/ML network, not just to (purportedly) "improving task distribution." Regardless, improvement of "task distribution" based on the specific factors of "changes in the one or more properties or the behavior of the agent and the tracked quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set" is an improvement of the "computer functionality itself." MPEP 2106.06(b) (citing McRO, Inc. v. Bandai Namco Games Am. Inc., as involving claims "directed to an improvement in computer-related technology" for "automatic lip synchronization and facial expression animation)”.
The Examiner respectfully disagrees.
With respect to the argument the Examiner notes that while the Applicant argues that Claim 1 is directed to improvements to the AI/ML network itself, and that the specific factors driving the claimed task distribution constitute an improvement to "computer functionality itself" analogous to McRO, is not persuasive. The Examiner respectfully maintains the rejection.
Applicant's reliance on McRO is misplaced because it fundamentally mischaracterizes what made those claims eligible. In McRO, the Federal Circuit found the claims patent eligible not merely because they used particular inputs to drive an output, but because the claims recited a specific set of rules specifically, rules for setting morph weights and transitions through phonemes that enabled the automation of a task that previously could only be performed subjectively by humans, and that produced a result in the technical field of computer animation that could not previously be achieved automatically. McRO. In other words, the claims in McRO recited the particular technical mechanism the specific rules themselves that achieved the improvement to computer animation technology. By contrast, Claim 1 here recites no analogous specific technical mechanism. The claim recites that the AI/ML network changes "an amount of interaction load" and "an amount of attention needed by the agent" based on "changes in the...properties or the behavior of the agent" and "tracked quality of performance," but nowhere does the claim specify the particular rules, algorithms, data structures, or computational mechanisms by which those changes are determined and effectuated. Claim 1 thus recites the desired outcome of the AI/ML network's adjustment a changed load balance without reciting the particular technical way in which that outcome is achieved. This is precisely the distinction the courts draw between claiming a particular solution (eligible, as in McRO) and merely claiming the idea of a solution or outcome (ineligible). See MPEP 2106.04(d)(1) and 2106.05(a); August 4, 2025 Memorandum at pp. 3-4.
Applicant's broader assertion that improvements to the AI/ML network's "knowledge base and capabilities" constitute an improvement to "computer functionality itself" similarly fails for the same reason addressed in the previous response: the specification describes these improvements only in generic terms, as the ML/AI layer being "trained to expand its knowledge base and capabilities to handle more types of interactions" using "data collected over time" (Spec. [0075]), without providing the technical detail that would be necessary to demonstrate that this constitutes an improvement to machine learning technology as a technical matter rather than an improvement to the abstract business task of automating more customer service interactions. An improvement in the capability to perform an abstract idea more broadly or more efficiently is not an improvement to technology. See Trading Technologies Int'l v. IBG, (claimed interface that provided traders more information improved the business process of market trading but did not improve computers or technology); MPEP 2106.05(a), subsection II. Furthermore, as noted in the August 4, 2025 Memorandum, in Recentive Analytics, Inc. v. Fox Corp., the Federal Circuit confirmed that steps incidental to automating an abstract idea even where machine learning is involved are not sufficient to confer eligibility, directly counseling against the proposition that the mere use of an AI/ML network to carry out an abstract business function, or to do so with improving proficiency over time, establishes a technological improvement. The claims here do not recite specific technical implementations that would demonstrate an improvement to computer technology or to machine learning as a technical field. The rejection is therefore maintained.
The Applicant argues on pages 16-17 that “Regardless, although "[a] claim may recite multiple judicial exceptions" "unless it is clear that a claim recites distinct exceptions, such as a law of nature and an abstract idea, care should be taken not to parse the claim into multiple exceptions, particularly in claims involving abstract ideas." MPEP 2106.04(II)(B). Even in claims that recite multiple abstract ideas that fall in the same or different groupings, "examiners should not parse the claim." Id. Pages 20-21 of Office Action sets forth precisely such improper parsing of the elements of Claim 1 into all three categories of abstract ideas, including by breaking elements apart in order to assign different putative abstract ideas to different parts of discrete elements. The Office Action even dismissively asserts that "functional language describing what is performed" (e.g., "iteratively performing... a system optimization operation") can be parsed out separately for purposes of then being completely disregarded by the Office Action.
For at least these reasons, Claim 1 and its dependent claims recite patent-eligible subject matter. For similar reasons, Claims 8 and 15 and their respective dependent claims recite patent-eligible subject matter”.
The Examiner respectfully disagrees.
With respect to the argument the Examiner notes that while the Applicant argues that the Office Action improperly parsed Claim 1 into multiple distinct abstract idea categories and particularly that "iteratively performing...a system optimization operation" was improperly disregarded as mere functional language is not persuasive. The Examiner respectfully maintains the rejection.
Applicant correctly states the governing principle from MPEP 2106.04, subsection II.B, that where a claim recites multiple abstract ideas, examiners should identify all relevant groupings for the record at Step 2A Prong One to make the analysis clear, but should then treat those multiple abstract ideas together as a single abstract idea for purposes of Step 2A Prong Two and Step 2B. That is precisely what the Office Action did. The rejection identified claim limitations that fell within the mathematical concepts, certain methods of organizing human activity, and mental processes groupings at Prong One as the MPEP expressly instructs, noting that "examiners should identify at least one abstract idea grouping, but preferably identify all groupings to the extent possible" and then expressly stated those limitations were "considered together as a single abstract idea" for the Prong Two analysis. See MPEP 2106.04(a)(2); MPEP 2106.04, subsection II.B. This is not improper parsing; it is the correct and thorough application of the USPTO's examination framework, mirroring the approach taken in the USPTO's own 2024 AI-SME Update Examples, which similarly identify limitations across multiple groupings (mathematical concepts and mental processes) and consolidate them for the Prong Two analysis. See 2024AISMEUpdateExamples4749, Example 47 (Claim 2), Example 48 (Claim 2).
Regarding Applicant's more specific objection that the Office Action improperly disregarded "iteratively performing...a system optimization operation" as mere functional language, this argument mischaracterizes the Office Action's analysis. The Office Action did not disregard those limitations it addressed them explicitly and in detail, including the sub-limitations of determining complexity, tracking quality of performance, changing interaction load, and expanding the knowledge base and capabilities of the AI/ML network. The characterization of "iteratively performing...a system optimization operation" as functional language was an observation that this overarching phrase describes what is performed at a high level of generality without itself imposing any specific technical constraint, which is relevant to the "apply it" and "idea of a solution versus particular solution" considerations under MPEP 2106.05(a) and 2106.05(f). Identifying functional language as such is a well-established aspect of the eligibility analysis, not a prohibited parsing technique. See August 4, 2025 Memorandum at pp. 3-4 (instructing examiners to consider "whether the claim recites only the idea of a solution or outcome").
Critically, even assuming arguendo that the Office Action had technically erred in how it identified the abstract idea groupings at Prong One, any such procedural error would be harmless because the dispositive question in this case is resolved at Step 2A Prong Two: whether the claim as a whole, evaluated as an ordered combination of all its limitations, integrates the recited judicial exception into a practical application. As addressed in detail in response to Arguments 2 and 3 above, it does not because the claim recites only functional outcomes (changing load, expanding capabilities) without the specific technical mechanisms, data structures, algorithms, or parameter adjustment details that would be necessary to demonstrate a genuine improvement to computer technology or machine learning technology, as required by MPEP 2106.05(a), Ex Parte Desjardins, and the line of Federal Circuit cases discussed therein. The identification of one abstract idea grouping versus three does not change this conclusion, because the claim fails at Prong Two regardless of how the abstract idea is characterized at Prong One. The rejection is therefore maintained.
The remaining Applicant's arguments filed 26 May 2026 have been fully considered but they are moot in view of new grounds of rejection as necessitated by amendment.
Claim Rejections - 35 USC 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-3, 5-10, 12-17, and 19-23 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter because the claim(s) 1-3, 5-10, 12-17, and 19-23 as a whole, considering all claim elements both individually and in combination, do not amount to significantly more than an abstract idea. The claim(s) 1-3, 5-10, 12-17, and 19-23 is/are directed to the abstract idea of distributing customer support between users based on contact load.
Step 1:
Regarding Step 1 of the Subject Matter Eligibility Test for Products and Processes, claims 1-7 and 21-23 are directed to a process/method, claim 15-19 is directed to a non-transitory machine-readable medium / manufacture, and claim 8-14 is directed to a system / machine. Therefore, the claims fall within the statutory categories of invention.
Step 2A Prong One:
The claims recite an abstract idea. Specifically, the independent claims 1, 8, and 15 recite the following limitations:
"obtaining...one or more properties of one or more interactions associated with an agent"
"determining...an amount of load for the agent generated by the one or more interactions based on the one or more properties of the one or more interactions"
"tracking...one or more changes in the one or more properties or a behavior of the agent"
"identifying...a new interaction based on analyzing agent performance data for an agent set including the agent"
“analyzing, suing an artificial intelligence/machine learning (AI/ML) network executing at least one processing device…”.
"determining, using a machine learning network, a complexity of the new interaction"
"tracking a quality of performance of the machine learning network"
"changing...an amount of interaction load for the machine learning network and an amount of attention needed by the agent"
"expanding a knowledge base and capabilities of the machine learning network to handle one or more additional types of interactions using data collected over time"
"determining one or more load scores for the agent based on the one or more properties of the one or more interactions, each load score associated with a measure of computing power and agent attention needed for the associated interaction"
Abstract Idea Grouping Analysis
Mathematical Concept:
These limitations recite mathematical calculations. Specifically, the limitations "determining an amount of load for the agent generated by the one or more interactions based on the one or more properties" and "determining one or more load scores for the agent based on the one or more properties of the one or more interactions" recite calculating numerical values (load amounts and scores) from input data through mathematical operations. The limitation "analyzing...an effect on agent performance...based on (i) the amount of load for the agent and (ii) the one or more changes" recites analyzing mathematical relationships between variables (load, property changes, behavior changes, and performance effects). The limitation "determining...a complexity of the new interaction" recites computing a complexity value through mathematical calculation. These are mathematical calculations and mathematical relationships that fall within the mathematical concepts grouping. See MPEP 2106.04(a)(2), subsection I.
Certain Methods of Organizing Human Activity:
These limitations fall within the "managing personal behavior or relationships or interactions between people" and "fundamental economic principles or practices" sub-groupings of certain methods of organizing human activity. Specifically, the claims recite managing how human customer service agents handle customer interactions by tracking agent workload, analyzing agent performance, determining agent capacity, and assigning new customer interactions to agents based on their workload and performance. These activities constitute managing human work activities and interpersonal interactions between agents and customers. The specification confirms this, describing "distribution of interactions between users based on load" and optimizing "how different agents are able to handle different interactions" to "maintain maximum overall performance of the agents" (Spec. Abstract, [0066]). See MPEP 2106.04(a)(2), subsection II; Versata Dev. Group, Inc. v. SAP Am., Inc.
Mental Process:
These limitations, under their broadest reasonable interpretation, cover performance of the limitations in the human mind. Specifically, the step of "tracking...one or more changes in the one or more properties or a behavior of the agent" encompasses mental observation where a supervisor could observe and mentally track changes in how an agent performs; the step of "analyzing...the effect of the new interaction" encompasses a mental evaluation where a supervisor could mentally analyze how workload affects performance through observation; the step of "identifying...a new interaction based on analyzing agent performance data" encompasses a mental judgment where a supervisor could mentally identify which interaction to assign next based on their knowledge of agent performance; the step of "determining...a complexity of the new interaction" encompasses a mental assessment where a supervisor could mentally evaluate how complex a customer interaction will be. The mere nominal recitation of "at least one processing device" and "machine learning network" does not take the claim limitations out of the mental processes grouping. See MPEP 2106.04(a)(2), subsection III; CyberSource Corp. v. Retail Decisions, Inc.
Because the claims recite limitations falling within multiple abstract idea groupings, these limitations are considered together as a single abstract idea. See MPEP 2106.04, subsection II.B. Accordingly, the claims recite judicial exceptions.
Step 2A Prong Two:
Identification of Additional Elements
The claims recite the following additional elements beyond the identified abstract idea:
"using at least one processing device of an electronic device" (claim 1)
“an artificial intelligence/machine learning network (claim 1, 8, and 15
"at least one processing device" (claim 8) / "at least one processor" (claim 15)
"electronic device" (claim 8)
"non-transitory machine-readable medium containing instructions" (claim 15)
"a machine learning network trained to handle at least one type of interaction on behalf of the agent or the agent set"
"iteratively performing...a system optimization operation" (functional language describing what is performed)
Analysis of Additional Elements
Improvement to Technology or Technical Field (MPEP 2106.05(a)):
The claims do not recite an improvement to the functioning of a computer or to any other technology or technical field. The specification describes the invention as addressing "distribution of interactions between users based on load" and improving "system performance" in managing agent workload, and describes the problem and solution in business terms: "distribution of interactions among agents can be different for various remote customer support environments" with a solution of "strategically distribute the interactions among the different agents" to "maintain maximum overall performance" (Spec. [0030]-[0034], [0088]). This describes using computers as tools to automate the business practice of assigning customer service work more efficiently, rather than improving computer functionality itself. The specification's description of system components (FIG. 1 showing standard electronic device elements; FIGS. 2-9 showing conventional operations like "operation 220," "operation 225," "operation 240") reveals only conventional computing elements performing conventional functions: obtaining data, performing calculations, tracking information, analyzing data, and distributing tasks.
The "machine learning network" is recited only in functional terms as being "trained to handle at least one type of interaction" and capable of "determining...a complexity" and "expanding a knowledge base." The specification describes the ML/AI layer generically as using "ML or AI algorithms, routines, or networks (such as one or more large language models, neural networks, and the like)" to handle "routine aspects of the interactions" (Spec. [0075]-[0076]). This describes using machine learning as a tool to automate routine customer service tasks, not any improvement to machine learning technology itself. Under the 2024 AI-SME Guidance, an improvement in the judicial exception itself (improving task distribution methods) is not an improvement in technology. See MPEP 2106.05(a); Trading Techs. Int'l, Inc. v. IBG LLC; SAP Am., Inc. v. InvestPic, LLC)
Particular Machine (MPEP 2106.05(b)):
The claims do not recite use of a particular machine that imposes meaningful limits on the claim. The recited "electronic device," "processing device," "at least one processor," and "machine learning network" are generic computing components described at a high level of generality without structural particularity. The machine learning network is described only by its generic capabilities, not by any specific architecture or configuration. The specification describes standard electronic device components (processor, memory, display, communication interface) without requiring any particular implementation (Spec. [0038]-[0048], FIG. 1). See MPEP 2106.05(b); Elec. Power Grp., LLC v. Alstom S.A.
Mere Instructions to Apply the Exception (MPEP 2106.05(f)):
The additional elements amount to no more than mere instructions to implement the abstract idea on a computer. The claims recite generic computing components (processing device, electronic device, machine learning network, non-transitory medium) performing generic computing functions (obtaining data, performing calculations, tracking information, analyzing data, determining complexity). Under the August 4, 2025 Memorandum on evaluating subject matter eligibility, the claims “recite only the idea of a solution or outcome” without details of how the solution is technically accomplished. For example, the claims recite “determining, using a machine learning network, a complexity of the new interaction” but provide no technical details about how this determination is made what algorithm is used, what features are extracted, what network architecture processes the data, or what makes this approach technically superior. This is tantamount to adding the words “apply it” to the judicial exception. See Alice Corp. Pty. Ltd. V. CLS Bank Int’l, 573 U.S. 208, 223 (2014); MPEP 2106.05(f).
Insignificant Extra-Solution Activity (MPEP 2106.05(g)):
The additional elements of "obtaining...one or more properties of one or more interactions" and "tracking...one or more changes in the one or more properties or a behavior of the agent" constitute insignificant extra-solution activity. The "obtaining" and "tracking" steps are mere data gathering that is necessary to provide inputs for the abstract idea calculations. These data gathering steps are incidental to the primary process of calculating load scores and assigning interactions, and do not impose any meaningful limits on practicing the abstract idea beyond requiring that data be collected. See MPEP 2106.05(g); OIP Techs., Inc. v. Amazon.com, Inc.
Considering the additional elements individually and in combination, the claim as a whole does not integrate the judicial exception into a practical application. The additional elements do not impose any meaningful limits on practicing the abstract idea. Accordingly, the claims are directed to an abstract idea.
Step 2B:
As discussed with respect to Step 2A Prong Two, the additional elements in the claims amount to no more than mere instructions to apply the exception using generic computer components and insignificant extra-solution activity (data gathering). The same analysis applies in Step 2B mere instructions to apply an exception using generic computer components and data gathering cannot provide an inventive concept. See MPEP 2106.05(f) and (g).
Well-Understood, Routine, Conventional Activity Analysis
The additional elements, when considered individually and in combination, are well-understood, routine, and conventional activities in the field. Specifically:
Using a processing device to perform calculations - The courts have recognized that using a computer processor to perform routine calculations is well-understood, routine, conventional activity. See MPEP 2106.05(d)(II), citing Parker v. Flook, (post-solution activity of adjusting values); Bancorp Servs., LLC v. Sun Life Assur. Co. of Canada (U.S.), ("The computer required by some of Bancorp's claims is employed only for its most basic function, the performance of repetitive calculations, and as such does not impose meaningful limits on the scope of those claims"); SAP Am., Inc. v. InvestPic, LLC,. The specification's description confirms conventional processing operations: "obtaining," "determining," "tracking," "analyzing," and "identifying" using generic electronic devices with conventional components (Spec. [0038]-[0048], FIG. 1).
Using machine learning networks - The specification itself provides evidence that machine learning networks are well-understood, routine, conventional. The specification describes using "ML or AI algorithms, routines, or networks (such as one or more large language models, neural networks, and the like)" (Spec. [0075]) without describing any particular innovations in these algorithms. This generic description indicates standard, commercially available machine learning technology. The specification describes the machine learning as performing conventional functions: processing customer interactions, handling routine conversations, determining complexity, and tracking performance (Spec. [0075]-[0076]). Under MPEP 2106.05(d), subsection III.A, a specification demonstrates the well-understood, routine, conventional nature of additional elements when it describes them "as a commercially available product, or in a manner that indicates that the additional elements are sufficiently well-known that the specification does not need to describe the particulars of such additional elements to satisfy 35 U.S.C. 112(a)." Additionally, the courts have recognized using conventional tools to automate processes as well-understood, routine, conventional. See In re TLI Commc'ns LLC Patent Litig., (using a conventional digital camera in a conventional manner).
Obtaining and tracking data - The courts have recognized receiving or collecting data as well-understood, routine, conventional activity. See MPEP 2106.05(d)(II), citing OIP Techs., Inc. v. Amazon.com, Inc., (gathering and presenting information); Elec. Power Grp., LLC v. Alstom S.A., (gathering data).
Performing calculations and determining scores - The courts have recognized performing calculations as well-understood, routine, conventional activity. See MPEP 2106.05(d)(II), citing Parker v. Flook, (mathematical calculations); Bancorp Servs., (performing calculations).
Assigning tasks based on calculated values - Assigning tasks to workers based on availability and workload is a fundamental business practice. The courts have recognized that automating fundamental business practices with computers is well-understood, routine, conventional. See Alice Corp., (automating fundamental business practices); Credit Acceptance Corp. v. Westlake Servs., (conventional business practice implemented on computer).
Considering the additional elements individually and in combination, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claims are not patent eligible.
Dependent Claims Analysis
The dependent claims do not add limitations that integrate the judicial exception into a practical application or provide an inventive concept:
Claims 2, 9, 16: These claims add steps of obtaining a new interaction from a queue, determining its load level, and selecting an agent based on load level with agents sorted by performance data. These limitations merely further specify the abstract idea of assigning work to available workers based on their capacity and current workload. Task queuing and worker selection based on availability and performance are fundamental business practices that constitute additional methods of organizing human activity and do not add any additional elements that would integrate the exception into a practical application or provide significantly more.
Claims 3, 10, 17: These claims specify that the properties of interactions include "a type of the interaction, data collected during the interaction, and one or more behaviors of a customer during the interaction." These limitations merely identify specific types of data inputs for the abstract idea calculations. Specifying particular data inputs constitutes additional insignificant extra-solution activity (data gathering) and does not integrate the abstract idea into a practical application or provide significantly more.
Claims 5, 12, 19: These claims add that the agent set includes at least one other agent and that performance data includes information for multiple agents. These limitations merely specify that the abstract idea is applied to a group of workers rather than a single worker, which does not change the nature of the abstract idea or add technological innovation.
Claims 6, 13, 20: These claims add optimizing the agent set based on load scores and system performance requirements. This further defines the abstract idea of managing worker allocation to meet business requirements a fundamental economic practice that does not integrate the exception into a practical application.
Claims 7, 14: These claims specify that the agent is a support agent. This is a field of use limitation that merely narrows the application of the abstract idea to a specific industry and does not impose any meaningful technical constraint.
Claim 21: This claim adds distributing the new interaction to the agent or another agent based on the load scores of the agents. This explicitly recites the abstract idea of assigning work based on worker capacity and does not add additional elements beyond the abstract idea.
Claim 22: This claim adds pausing the new interaction until a suitable agent is available. Placing tasks in a queue when no worker is available is a conventional business practice that constitutes well-understood, routine, conventional activity and does not provide an inventive concept.
Claim 23: This claim adds creating a pre-sorted queue of agents based on capability information for each agent. Maintaining a sorted list of workers based on their skills and capabilities is a conventional management practice. Implementing this using computer data structures (a queue) is well-understood, routine, conventional activity.
For the foregoing reasons, claims 1, 2, 3, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 19, 20, 21, 22, and 23 are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent may not be obtained through the invention is not identically disclosed or described as set forth in section 102, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim 1-3, 5-10, 12-17, and 19-23 is/are rejected under 35 U.S.C. 103 as being unpatentable over Barinov et al. (U.S. Patent Publication 2020/0280635 A1) (hereafter Barinov) in view of McCormack et al. (U.S. Patent Publication 2015/0103995 A1) (hereafter McCormack).
Referring to Claim 1, Barinov teaches a method comprising:
obtaining, using at least one processing device of an electronic device, one or more properties of one or more interactions associated with an agent (see; par. [0044] and par. [0052] of Barinov teaches The interaction server a processing device obtains the properties of each interaction associated with an agent: specifically, the media type and the predicted difficulty/complexity classification. These are "one or more properties of one or more interactions associated with an agent" as claimed. The system captures these properties as inputs before computing each interaction's cost value and processing device (interaction server 160) explicitly receives and acts on the interaction's properties (media type and difficulty classification) directly teaching obtaining properties of interactions associated with an agent).
determining, using the at least one processing device, an amount of load for the agent generated by the one or more interactions associated with the agent, wherein the amount of load is based on the one or more properties of the one or more interactions (see; par. [0020] of Barinov teaches The agent's load (the "used capacity value") is explicitly determined by summing the cost values of all concurrent interactions. Each cost value is derived from the interaction's properties (media type + difficulty). Load is therefore directly determined based on the properties of the interactions, par. [0060] the step of determining the agent's load (used capacity) by summing per-interaction cost values derived from interaction properties precisely teaching”.
tracking, using the at least one processing device, one or more changes in the one or more properties or a behavior of the agent regarding the one or more interactions (see; par. [0057] of Barinov teaches This single paragraph teaches all aspects of the tracking limitation: (1) tracking changes in interaction properties (media type escalating from chat to voice; difficulty reclassified); (2) tracking changes in agent behavior (agent modifies difficulty rating; agent updates own capacity). The stat server continuously monitors and recalculates based on these tracked changes, par. [0049] property changes are tracked in real time during operation the system supports continuous runtime monitoring of changes in interaction properties and agent behavior.
identifying, using the at least one processing device, a new interaction (see; par. [0049] of Barinov teaches the processing device (SIP server/interaction server) identifies a new incoming interaction as the event that initiates the entire routing process directly teaching "identifying…a new interaction, par. [0052] identified new interaction is processed by the interaction server confirming the processing device identifies and receives the new interaction before any routing decision is made).
based on analyzing, using an artificial intelligence/machine learning (AI/ML) network executing on the at least one processing device, an effect of the new interaction on agent performance for the agent by analyzing agent performance data for an agent set including the agent, wherein the effect of the new interaction on agent performance for the agent is analyzed based on (i) the amount of load for the agent and (ii) the one or more properties or the behavior of the agent (see; par. [0020] of Barinov teaches that ML is applied to optimize the cost and capacity values i.e., an AI/ML network analyzes and optimizes how interaction costs and agent capacities are set, directly teaching that an AI/ML network is used to analyze the effect of interactions on agent performance, par. [0059] the new interaction on agent performance is analyzed using (i) the agent's current load (used capacity sum of cost values of ongoing interactions) and (ii) the agent's performance data (experience level, training, recent performance characteristics) and interaction properties (cost values weighted by attention span). Both prongs of the claimed dual-basis analysis are explicitly present, par. [0054] routing logic analyzes agent performance data (available/used capacity, experience, special abilities) across the full agent set teaching that the effect of the new interaction is analyzed based on agent performance data for an agent set including the agent).
iteratively performing, using the at least one processing device, a system optimization operation, wherein iteratively performing the system optimization operation comprises (see; par. [0049] of Barinov teaches continuously and iteratively runs its optimization cost/capacity values update dynamically in real time, confirming an iteratively performed system optimization operation, par. [0055] interaction assignment the stat server recomputes updated capacity values and sends them back to the routing server for use in subsequent routing decisions an explicit iterative optimization loop executing continuously as new interactions arrive, and par. [0056] the iterative loop completes and resets upon interaction termination the stat server again recomputes and feeds updated values to the routing server, confirming the continuous, iterative nature of the system optimization operation).
wherein determining the amount of load for the agent comprises determining one or more load scores for the agent based on the one or more properties of the one or more interactions, each load score associated with a measure of computing power for the AI/ML network and agent attention needed for the associated interaction (see; par. [0059] of Barinov teaches each interaction's cost value (i.e. load score) is explicitly weighted to reflect the agent attention span required for that interaction type based on the interaction's properties. This teaches "load scores… associated with… agent attention needed for the associated interaction”.
Barinov does not explicitly disclose the following limitations, however,
McCormack teaches determining, using a the AI/ML network, a complexity of the new interaction, the AI/ML network trained to handle at least one type of interaction on behalf of the agent or the agent set (see; par. [0039] of McCormack teaches automated customer engagement server is an AI/ML-based NLP system that is trained to handle customer chat interactions entirely on behalf of live agents directly teaching "the AI/ML network trained to handle at least one type of interaction on behalf of the agent or the agent set.", par. [0040] the confidence factor generated by the AI/ML automated engagement server is a direct, quantified measure of the complexity of each interaction as assessed by the AI/ML network a low confidence factor means the interaction is too complex for the AI/ML to handle adequately. This directly teaches "determining… a complexity of the new interaction" using the AI/ML network”, par. [0041] the confidence factor (complexity measure) has a dynamic range and is recalculated throughout the engagement based on accumulated session data confirming the AI/ML network continuously determines the complexity of each interaction as it evolve), and
tracking a quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set (see; par. [0055] of McCormack the engine repeatedly monitors and updates the confidence factor (quality of performance of the AI/ML system) throughout each engagement confirming iterative, ongoing tracking of the AI/ML network's quality of performance as it handles interactions on behalf of agents), and
changing, based on the tracked one or more changes in the one or more properties or the behavior of the agent and the tracked quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, an amount of interaction load for the machine learning AI/ML network and an amount of attention needed by the agent for the at least one type of interaction (see; par. [0046] of McCormack teaches When the AI/ML performs well (high confidence = low complexity), it continues handling more interactions and the live agent's attention requirement is reduced (AI load high, agent attention low). When AI/ML performance degrades (low confidence = high complexity), the threshold shifts, routing more interactions to live agents (AI load reduced, agent attention increased). This dynamic adjustment driven by tracked AI/ML quality (confidence factor) and tracked agent behavior/availability directly teaches "changing…an amount of interaction load for the AI/ML network and an amount of attention needed by the agent…based on tracked changes in agent behavior and tracked quality of performance of the AI/ML network."
par. [0052] when AI/ML quality drops (low confidence), agent attention load is explicitly increased by bringing in a live agent the AI/ML's interaction load decreases and agent attention needed increases. This is a concrete example of changing the amount of interaction load for the AI/ML network and the amount of attention needed by the agent, and par. [0054] AI/ML quality is high (high confidence = low complexity), the AI/ML handles the full interaction load without agent involvement agent attention needed is zero for that interaction. This is the converse teaching: AI/ML load remains high and agent attention remains low when performance quality is good).
expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time as the agent set and the AI/ML network handle interactions, wherein the data collected over time includes the tracked quality of performance, and wherein expanding the knowledge base and the capabilities of the AI/ML network comprises determining to leave the new interaction for the agent to handle based on the complexity of the new interaction (see; par. [0057] of McCormack teaches live agents collaborating and the system continuously collecting confidence factor data (tracked quality of performance) from each engagement, par. [0052] when the interaction's complexity exceeds the AI/ML's capability (low confidence factor), the system routes the interaction to the live agent i.e., it determines to leave the interaction for the agent to handle based on the complexity of the new interaction, par. [0054] The inverse case when complexity is low (high confidence), the system determines NOT to leave the interaction for the agent; the AI/ML handles it. Together with [0052], this paragraph pair teaches that the routing decision between AI/ML and agent is made based entirely on the measured complexity (confidence factor)).
The Examiner notes that Barinov similar to the instant application teaches extended agent capacity for routing interactions. Specifically, Barinov discloses routing process may include receiving the first interaction, the classifier, classifying the interaction by type and determining cost value of the interaction is therefore viewed as analogous art in the same field of endeavor. Additionally, McCormack teaches prioritizing agent intervention into automated customer engagements and as it is comparable in certain respects to Barinov which management of conversations in a contact center as well as the instant application it is viewed as analogous art and is viewed as reasonably pertinent to the problem faced by the inventor. This provides support that it would be obvious to combine the references to provide an obviousness rejection.
Barinov discloses the routing process may include receiving the first interaction, the classifier, classifying the interaction by type and determining cost value of the interaction. However, Barinov fails to disclose determining, using a the AI/ML network, a complexity of the new interaction, the AI/ML network trained to handle at least one type of interaction on behalf of the agent or the agent set, tracking a quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, changing, based on the tracked one or more changes in the one or more properties or the behavior of the agent and the tracked quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, an amount of interaction load for the machine learning AI/ML network and an amount of attention needed by the agent for the at least one type of interaction, and expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time as the agent set and the AI/ML network handle interactions, wherein the data collected over time includes the tracked quality of performance, and wherein expanding the knowledge base and the capabilities of the AI/ML network comprises determining to leave the new interaction for the agent to handle based on the complexity of the new interaction.
McCormack discloses determining, using a the AI/ML network, a complexity of the new interaction, the AI/ML network trained to handle at least one type of interaction on behalf of the agent or the agent set, tracking a quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, changing, based on the tracked one or more changes in the one or more properties or the behavior of the agent and the tracked quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, an amount of interaction load for the machine learning AI/ML network and an amount of attention needed by the agent for the at least one type of interaction, and expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time as the agent set and the AI/ML network handle interactions, wherein the data collected over time includes the tracked quality of performance, and wherein expanding the knowledge base and the capabilities of the AI/ML network comprises determining to leave the new interaction for the agent to handle based on the complexity of the new interaction.
It would be obvious to one of ordinary skill in the art to include in the task management
(system/method/apparatus) of Barinov determining, using a the AI/ML network, a complexity of the new interaction, the AI/ML network trained to handle at least one type of interaction on behalf of the agent or the agent set, tracking a quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, changing, based on the tracked one or more changes in the one or more properties or the behavior of the agent and the tracked quality of performance of the AI/ML network in handling the at least one type of interaction on behalf of the agent or the agent set, an amount of interaction load for the machine learning AI/ML network and an amount of attention needed by the agent for the at least one type of interaction, and expanding a knowledge base and capabilities of the AI/ML network to handle one or more additional types of interactions using data collected over time as the agent set and the AI/ML network handle interactions, wherein the data collected over time includes the tracked quality of performance, and wherein expanding the knowledge base and the capabilities of the AI/ML network comprises determining to leave the new interaction for the agent to handle based on the complexity of the new interaction as taught by McCormack since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Additionally, Barinov and McCormack teach the collecting and analysis of data in order to maximize the performance of call center individuals based on performance and load balancing and they do not contradict or diminish the other alone or when combined.
Referring to Claim 2, see discussion of claim 1 above, while Barinov in view of McCormack teaches the method above, Barinov further discloses a method having the limitations of,
obtaining the new interaction from a queue of interactions to assign and determining a load level of the new interaction; and selecting the agent from a group of available agents in the agent set based on the load level of the new interaction, the group of available agents being sorted based on the agent performance data (see; par. [0053] of Barinov teaches the new interaction arrives from the SIP queue par. [0051], its load level (cost value) is determined par. [0052], and the stat server composes a filtered list of available agents sorted by their capacity/performance data directly teaching obtaining the new interaction from a queue, determining its load level, and selecting from a sorted group of available agents based on agent performance data, and par. [0054] the routing server selects the agent from the sorted list based on the new interaction's load level (cost value) compared against agents' available capacity (agent performance data).
Referring to Claim 3, see discussion of claim 1 above, while Barinov in view of McCormack teaches the method above, Barinov further discloses a method having the limitations of,
the one or more properties of the one or more interactions include at least one of: a type of the interaction, data collected during the interaction, and one or more behaviors of a customer during the interaction (see; par. [0044] of Barinov teaches a media type is explicitly "a type of the interaction." Difficulty prediction (taught in [0048] based on customer name and stated intent) is "data collected during the interaction." Both types of interaction properties are expressly disclosed, par. [0048] where a customer name and stated intent data collected from/about the customer at the time of the interaction are used as interaction properties to predict difficulty. This teaches "data collected during the interaction" as a property of the interaction).
Referring to Claim 5, see discussion of claim 1 above, while Barinov in view of McCormack teaches the method above, Barinov further discloses a method having the limitations of,
the agent set further includes at least one other agent; and the agent performance data for the agent set includes information associated with agent performance for the at least one other agent (see; par. [0022] of Barinov teaches operating with a plurality of agent devices establishing a multi-agent pool (agent set) that includes multiple agents, par. [0054] the routing logic uses performance data (available/used capacity) for each agent in the pool teaching that agent performance data for the agent set includes information for every agent in the set, including agents other than the specific agent).
Referring to Claim 6, see discussion of claim 5 above, while Barinov in view of McCormack teaches the method above, Barinov further discloses a method having the limitations of,
optimizing the agent set based on a load score for each agent and one or more system performance requirements (see; par. [0054] of Barinov teaches he routing/optimization process uses load scores (capacity values) combined with additional system performance requirements (experience level, special abilities) to optimize which agent receives each interaction teaching optimization based on load scores and system performance requirements, and par. [0059] agent capability ratings (experience, training, recent performance) are system performance requirements used in conjunction with load scores to optimize agent assignments).
Referring to Claim 7, see discussion of claim 1 above, while Barinov in view of McCormack teaches the method above, Barinov further discloses a method having the limitations of,
the agent is a support agent (see; par. [0023] of Barinov teaches explicitly customer service and support agents handling customer service, help desk, and related support functions they are definitionally (i.e. support agents).
Referring to Claim 8, Hill in view of Barinov in view of McCormack teaches an electronic device. Claim 8 recites the same or similar limitations as those addressed above in claim 1, Claim 8 is therefore rejected for the same reasons as set forth above in claim 1, except for the following noted exception.
at least one processing device (see; par. [0010] of Barinov teaches a hardware processor).
Referring to Claim 9, see discussion of claim 8 above, while Barinov in view of McCormack teaches the electronic device above Claim 9 recites the same or similar limitations as those addressed above in claim 2, Claim 9 is therefore rejected for the same or similar limitations as set forth above in claim 2.
Referring to Claim 10, see discussion of claim 8 above, while Barinov in view of McCormack teaches the electronic device above Claim 10 recites the same or similar limitations as those addressed above in claim 3, Claim 10 is therefore rejected for the same or similar limitations as set forth above in claim 3.
Referring to Claim 12, see discussion of claim 8 above, while Barinov in view of McCormack teaches the electronic device above Claim 12 recites the same or similar limitations as those addressed above in claim 5, Claim 12 is therefore rejected for the same or similar limitations as set forth above in claim 5.
Referring to Claim 13, see discussion of claim 12 above, while Barinov in view of McCormack teaches the electronic device above Claim 13 recites the same or similar limitations as those addressed above in claim 6, Claim 13 is therefore rejected for the same or similar limitations as set forth above in claim 6.
Referring to Claim 14, see discussion of claim 8 above, while Barinov in view of McCormack teaches the electronic device above Claim 14 recites the same or similar limitations as those addressed above in claim 7, Claim 14 is therefore rejected for the same or similar limitations as set forth above in claim 7.
Referring to Claim 15, Barinov in view of McCormack teaches a non-transitory machine-readable medium. Claim 15 recites the same or similar limitations as those addressed above in claim 1, Claim 15 is therefore rejected for the same reasons as set forth above in claim 1.
Referring to Claim 16, see discussion of claim 15 above, while Barinov in view of McCormack teaches the non-transitory machine-readable medium above Claim16 recites the same or similar limitations as those addressed above in claim 2, Claim 16 is therefore rejected for the same or similar limitations as set forth above in claim 2.
Referring to Claim 17, see discussion of claim 15 above, while Barinov in view of McCormack teaches the non-transitory machine-readable medium above Claim17 recites the same or similar limitations as those addressed above in claim 3, Claim 17 is therefore rejected for the same or similar limitations as set forth above in claim 3.
Referring to Claim 19, see discussion of claim 15 above, while Barinov in view of McCormack teaches the non-transitory machine-readable medium above Claim19 recites the same or similar limitations as those addressed above in claim 5, Claim 19 is therefore rejected for the same or similar limitations as set forth above in claim 5.
Referring to Claim 20, see discussion of claim 19 above, while Barinov in view of McCormack teaches the non-transitory machine-readable medium above Claim 20 recites the same or similar limitations as those addressed above in claim 6, Claim 20 is therefore rejected for the same or similar limitations as set forth above in claim 6.
Referring to Claim 21, see discussion of claim 1 above, while Barinov in view of McCormack teaches the method above, Barinov further disclose a method having the limitations of, however,
the agent set further includes at least one other agent; and the method further comprises distributing the new interaction to the agent or the at least one other agent based on the load scores of the agents (see; par. [0054] of Barinov teaches each new interaction to a selected agent from the multi-agent pool based directly on load scores (available/used capacity values) the agent with the best load score relative to the new interaction's cost value is selected from the agent set).
Referring to Claim 22, see discussion of claim 21 above, while Barinov in view of McCormack teaches the method above, Barinov further disclose a method having the limitations of, however,
distributing the new interaction comprises pausing the new interaction until a suitable agent is available (see; par. [0047] of Barinov teaches when no agent has sufficient available capacity, routing is explicitly prevented the interaction remains in the queue/workbin until an agent with adequate capacity becomes available. This teaches pausing the new interaction until a suitable agent is available, and par. [0034] the workbin is the explicit queuing/pausing mechanism interactions await assignment in workbins until a suitable agent with sufficient capacity is available.).
Referring to Claim 23, see discussion of claim 21 above, while Barinov in view of McCormack teaches the method above, Barinov further disclose a method having the limitations of, however,
creating a pre-sorted queue of agents based on capability information for each agent (see; par. [0053] of Barinov teaches the stat server composes (creates) a filtered, ordered list of agents based on each agent's capacity information this pre-composed ranked list is a "pre-sorted queue of agents based on capability information for each agent.", and par. [0059] the pre-sorted queue is built from each agent's capability information (experience level, training, recent performance) directly teaching "a pre-sorted queue of agents based on capability information for each agent."
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Munoz et al. (U.S. Patent 12,425,519 B2) discloses systems and methods for relative gain in predictive routing.
Kassel (U.S. Patent 11,968,329 B2) discloses systems and methods relating to routing incoming interactions in a contact center.
Saushkin (U.S. Patent 10,477,026 B2) discloses a system for routing interactions using bio-performance attributes of persons as dynamic input.
Surendra et al. (U.S. Patent 9,805,330 B2) discloses a system and method for root cause analysis and early warning of inventory problems.
Kamath et al. (U.S. Patent Publication 2017/0364847 A1) discloses a system and method for solving large scale supply chain planning problems with integer constraints.
Bellini et al. (U.S. Patent 5,974,395) discloses a system and method for extended enterprise planning across a supply chain.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to STEPHEN S SWARTZ whose telephone number is (571)270-7789. The examiner can normally be reached Mon-Fri 9:00 - 6:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Boswell Beth can be reached at 571 272-6737. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/S.S.S/Examiner, Art Unit 3625
/DYLAN C WHITE/Primary Examiner, Art Unit 3625
Prosecution Insights