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 .
Response to Amendment
The amendment filed on September 26th, 2025 has been entered and Claims 1-20 is/are pending. Applicant’s amendments to the Claims have overcome the 35 U.S.C 101 rejection(s) previously set forth in the Non-Final Action mailed August 11th, 2025.
Response to Arguments
Applicant's arguments filed September 26th, 2025 have been fully considered but they are not persuasive.
Applicant’s arguments with respect to claim(s) 1, 10, and 19 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 10, and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yang et al., Non-Patent Literature (“On hyperparameter optimization of machine learning algorithms: Theory and practice”) in view of Schwartz et al., Non-Patent Literature (“Green AI.”) and Justus et al., Non-Patent Literature (“Predicting the Computational Cost of Deep Learning Models”)
Regarding claim 1 and analogous claim 10, and 19:
Yang teaches:
receiving, by a device, a configuration associated with a machine learning model; (Introduction, paragraph 1, “named hyper-parameters, cannot be directly estimated from data learning and must be set before training a ML model because they define the model architecture [4]. Hyper-parameters are the parameters that are used to either configure a ML model (e.g., the penalty parameter C in a support vector machine, and the learning rate to train a neural network).”)
receiving, by the device, a first hyperparameter set associated with the machine learning model; (introduction, paragraph 3, “The main aim of HPO is to automate hyper-parameter tuning process and make it possible for users to apply machine learning models to practical problems effectively [3]. The optimal model architecture of a ML model is expected to be obtained after a HPO process. Some important reasons for applying HPO techniques to ML models are as follows [6]: 1. It reduces the human effort required, since many ML developers spend considerable time tuning the hyper-parameters, especially for large datasets or complex ML algorithms with a large number of hyper-parameters. 2. It improves the performance of ML models. Many ML hyperparameters have different optimums to achieve best performance in different datasets or problems. 3. It makes the models and research more reproducible. Only when the same level of hyper-parameter tuning process is implemented can different ML algorithms be compared fairly; hence, using a same HPO method on different ML algorithms also helps to determine the most suitable ML model for a specific problem.”)
Yang does not explicitly teach:
estimating, by the device, a first quantity of floating point operations (FLOPs) associated with one or more epochs, for the machine learning model, based on the first hyperparameter set; and
outputting, to a user, an indication of a first energy consumption associated with training the machine learning model based on the first quantity of FLOPs.
wherein estimating the first quantity of the FLOPs comprises
identifying, by the device, a quantity of multiply-and-accumulate (MAC) operations
based on an architecture of the machine learning model and a selected optimization algorithm for the machine learning model
identifying a quantity of layers through which a training data set passes, and applying activation functions and weights to each layer of the quantity of the layers to estimate the first quantity of the FLOPs
Schwartz teaches:
estimating, by the device, a first quantity of floating point operations (FLOPs) associated with one or more epochs, for the machine learning model, based on the first hyperparameter set; and (Page 60, col 1-2, “First, it directly computes the amount of work done by the running machine when executing a specific instance of a model and is thus tied to the amount of energy consumed. Second, FPO is agnostic to the hardware on which the model is run. This facilitates fair comparisons between different approaches, unlike the measures described above. Third, FPO is often correlated with the running time of the model5 (though see discussion below). Unlike asymptotic runtime, FPO also considers the amount of work done at each time step.”)
outputting, to a user, an indication of a first energy consumption associated with training the machine learning model based on the first quantity of FLOPs. (Page 60, Col 1, paragraph 2, “FPO. As a concrete measure, we suggest reporting the total number of floating-point operations (FPO) required to generate a result. FPO provides an estimate of the amount of work performed by a computational process. It is computed analytically by defining a cost to two base operations, ADD and MUL. Based on these operations, the FPO cost of any machine learning abstract operation (for example, a tanh operation, a matrix multiplication, a convolution operation, or the BERT model) can be computed as a recursive function of these two operations. FPO has been used in the past to quantify the energy footprint of a model”)
Schwartz and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Schwartz it would have been obvious for a person of ordinary skill in the art to apply the teachings of Schwartz to Yang before the effective filing date of the claimed invention in order to improve the efficiency of energy consumption for training a machine learning model (Schwartz, Page 59, Col 1, “The term Green AI refers to AI research that yields novel results while taking into account the computational cost, encouraging a reduction in resources spent. Whereas Red AI has resulted in rapidly escalating computational (and thus carbon) costs, Green AI promotes approaches that have favorable performance/efficiency trade-offs. If measures of efficiency are widely accepted as important evaluation metrics for research alongside accuracy, then researchers will have the option of focusing on the efficiency of their models with positive impact on both inclusiveness and the environment.”)
Justus teaches:
wherein estimating the first quantity of the FLOPs comprises (Section III, paragraph 1, “The execution time required during a forward pass through a neural network is bounded from below by the number of floating point operations (FLOPs) [9]. This FLOP count depends on the deep neural network architecture and the amount of data (i.e., wherein the FLOP count is interpreted as the first quantity of the FLOPs)”)
identifying, by the device, a quantity of multiply-and-accumulate (MAC) operations (Page 3874, paragraph 2, “As convolutions and fully connected layers (vector-matrix multiplications) make up the most significant part of the execution time in the majority of deep neural networks, we focus on these types of operations (i.e., wherein the vector-matrix multiplications is interpreted to include multiply-and-accumulate (MAC) operations)”)
based on an architecture of the machine learning model and a selected optimization algorithm for the machine learning model (Section III, paragraph 1, “The execution time required during a forward pass through a neural network is bounded from below by the number of floating point operations (FLOPs) [9]. This FLOP count depends on the deep neural network architecture [architecture of the machine learning model]”…(Section IV, paragraph 1, “We define here the features which could influence the prediction of execution times when performing training”…“Optimiser used for locating a minima within the loss function space. These can include Gradient Descent, Adadelta, Adagrad, Momentum, Adam and RMS Prop (i.e., wherein the optimization algorithm are listed)”)
identifying a quantity of layers through which a training data set passes, and (Section III, paragraph 1, “The execution time required during a forward pass through a neural network is bounded from below by the number of floating point operations (FLOPs) [9]. This FLOP count depends on the deep neural network architecture (i.e., wherein ‘during a forward pass’ under broadest reasonable interpretation (BRI) includes input data (i.e., training data set passes))”…(Section IV, A, “These relate to those features of a particular layer within the neural network [layers] and in particular to the hyperparameters related to that layer… It should be noted that each individual layer within the network may possess different values for these features.”
applying activation functions and weights to each layer of the quantity of the layers to estimate the first quantity of the FLOPs (Section III, paragraph 1, “The execution time required during a forward pass through a neural network is bounded from below by the number of floating point operations (FLOPs) [9]. This FLOP count depends on the deep neural network architecture (i.e., wherein the execution time is based on the number of FLOPs which further depends on the dnn architecture)”…(Section IV, A, “These relate to those features of a particular layer within the neural network [layers] and in particular to the hyperparameters related to that layer”…“ Activation function used on the individual neurone. These can include none, ReLU, softmax, sigmoid and tanh. They can be encoded into the feature set using one-hot encoding”…“ As convolutions and fully connected layers (vector-matrix multiplications) make up the most significant part of the execution time in the majority of deep neural networks (i.e., wherein vector-matrix multiplications are interpreted to include weights)”
Justus and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Justus it would have been obvious for a person of ordinary skill in the art to apply the teachings of Justus to Yang before the effective filing date of the claimed invention in order to improve the efficiency of energy consumption for training a machine learning model (Justus, Abstract,“ Deep learning is rapidly becoming a go-to tool for many artificial intelligence problems due to its ability to outperform other approaches and even humans at many problems. Despite its popularity we are still unable to accurately predict the time it will take to train a deep learning network to solve a given problem. This training time can be seen as the product of the training time per epoch and the number of epochs which need to be performed to reach the desired level of accuracy.”)
Claim(s) 2, and 11 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz and Justus, in view of Lacoste et al., Non-Patent Literature (“Quantifying the Carbon Emissions of Machine Learning.”)
Regarding claim 2 and analogous claim 11:
Yang, as modified by Schwartz and Justus, teaches the method of claim 1.
Yang, as modified by Schwartz and Justus, does not explicitly teach:
receiving, by the device, an indication of hardware to be used for training the machine learning model; and
determining, by the device, the first energy consumption associated with training the machine learning model based on a thermal design power (TDP) associated with the hardware.
Lacoste teaches:
receiving, by the device, an indication of hardware to be used for training the machine learning model; and (Section 3, second paragraph, “This tool, currently in its alpha version, takes as input the details regarding the training of an ML model: the geographical zone of the server, the type of GPU, and the training time, and gives as output the approximate amount of CO2eq produced.”)
determining, by the device, the first energy consumption associated with training the machine learning model based on a thermal design power (TDP) associated with the hardware. (Page 4, paragraph 2, “To perform a comparison between different devices, their compute efficiency can be estimated in FLOPS/W. This estimation is based on their theoretical peak performance with respect to their Thermal Design Power (TDP). Using this approach, it can be found that CPUs can be 10 times less efficient than GPUs while TPU 3 can be 4 to 8 times more efficient than GPUs [23] (refer to Table 4 for details). Interestingly, in contexts where low power consumption and efficiency are important, e.g., for embedded applications, GPUs such as the Jetson AGX Xavier can be 10 to 20 times more efficient than traditional GPUs.”)
Lacoste and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Lacoste it would have been obvious for a person of ordinary skill in the art to apply the teachings of Lacoste to Yang before the effective filing date of the claimed invention to apply estimations using floating operation points (FLOPs) and thermal design power (TDP) in order to improve the efficiency of energy consumption for training a machine learning model (Lacoste, Introduction, paragraph 1, “Furthermore, new models often need to beat existing challenges, which entails training on more GPUs, with larger datasets, for a longer time. This expansion brings with it ever-growing costs in terms of the energy needed to fuel it. This trend has been the subject of recent studies aiming to evaluate the climate impact of AI, which have predominantly put the focus on the environmental cost of training large-scale models.”)
Claim(s) 3, and 12 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz and Justus, further in view of Luo et al., Non-Patent Literature (“A review of automatic selection methods for machine learning algorithms and hyper-parameter values.”)
Regarding claim 3 and analogous claim 12:
Yang, as modified by Schwartz and Justus, teaches the method of claim 1.
Yang, as modified by Schwartz and Justus, does not explicitly teach:
outputting, to the user, an indication of a recommended optimization algorithm for the machine learning model
wherein the configuration associated with the machine learning model includes an optimization algorithm selected by the user
Luo teaches:
outputting, to the user, an indication of a recommended optimization algorithm for the machine learning model (Section 1.3, Col 2, “Using an automatic selection method, the user of a machine learning software tool can skip the manual and iterative process of selecting an effective algorithm and/or combination of hyper-parameter values, which is labor intensive and requires a high skill set in machine learning” and (Section 1.3, “To make machine learning accessible to layman users, computer science researchers have proposed various automatic selection methods for machine learning algorithms and/or hyper-parameter values for a given supervised machine learning problem,” i.e., wherein an effective algorithm is recommended and selected.)
wherein the configuration associated with the machine learning model includes an optimization algorithm selected by the user (Section 1.3, Col 2, “Using an automatic selection method, the user of a machine learning software tool can skip the manual and iterative process of selecting an effective algorithm and/or combination of hyper-parameter values, which is labor intensive and requires a high skill set in machine learning,” i.e., wherein the optimizer algorithm is manually selected by the user.)
A person of ordinary skill in the art would reasonably find the teachings of Luo to be helpful in solving the problem of estimating energy consumption for machine learning models in Yang. In view of the teachings of Luo it would have been obvious for a person of ordinary skill in the art to apply the teachings of Luo to Yang before the effective filing date of the claimed invention to combine recommending an optimized algorithm selected by the user in order to improve the efficiency of energy consumption for training a machine learning model (Luo, Abstract, “Before a machine learning model is trained, the user of a machine learning software tool typically must manually select a machine learning algorithm and set one or more model parameters termed hyper-parameters. The algorithm and hyper-parameter values used can greatly impact the resulting model’s performance.”)
Claim(s) 4, and 13 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz and Justus, further in view of Desislavov et al., Non-Patent Literature (“Compute and Energy Consumption Trends in Deep Learning Inference.”)
Regarding claim 4 and analogous claim 13:
Yang, as modified by Schwartz and Justus, teaches the method of claim 1.
Yang, as modified by Schwartz and Justus, does not explicitly teach:
estimating, by the device, a second quantity of FLOPs associated with the one or more epochs, for the machine learning model, based on a second hyperparameter set; and
outputting, to the user, an indication of a second energy consumption associated with training the machine learning model based on the second quantity of FLOPs.
Desislavov teaches:
estimating, by the device, a second quantity of FLOPs associated with the one or more epochs, for the machine learning model, based on a second hyperparameter set; and (Page 2, paragraph 5, “We focus our analysis on inference FLOPs (Floating Point Operations) required to process one input item (image or text fragment)”, and “Having all this data we finally estimate how much energy is needed to perform one inference step with a given DNN.”)
outputting, to the user, an indication of a second energy consumption associated with training the machine learning model based on the second quantity of FLOPs (Page 9, paragraph 2, “Once we have estimated the inference FLOPs for a range of models and the GFLOPS per Watt for different GPUs, we can estimate the energy (in Joules) consumed in one inference. We do this by dividing the FLOPs for the model by FLOPS per Watt for the GPU” and,” They show that energy and FLOPs are highly correlated for the same architecture, but the correlation decreases when different architectures are mixed.”)
A person of ordinary skill in the art would reasonably find the teachings of Desislavov to be helpful in solving the problem of estimating energy consumption for machine learning models in Yang. In view of the teachings of Desislavov it would have been obvious for a person of ordinary skill in the art to apply the teachings of Desislavov to Yang before the effective filing date of the claimed invention in order to improve the efficiency of energy consumption for training a machine learning model (Desislavov, Abstract, “we account for more specific and powerful hardware (leading to higher FLOPS) that is usually accompanied with important energy efficiency optimisations. We also move the focus from the first implementation of a breakthrough paper towards the consolidated version of the techniques one or two year later. Under this distinctive and comprehensive perspective, we study relevant models in the areas of computer vision and natural language processing: for a sustained increase in performance, we see a much softer growth in energy consumption than previously anticipated.”)
Claim(s) 5, and 14 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz, Justus and Desislavov, further in view of Strubell et al., Non-Patent Literature (“Energy and Policy Considerations for Deep Learning in NLP.”)
Regarding claim 5 and analogous claim 14:
Yang, as modified by Schwartz, Justus and Desislavov, teaches the method of claim 4.
Yang, as modified by Schwartz, Justus and Desislavov, does not explicitly teach:
outputting a visual graph of the first energy consumption and the second energy consumption relative to the first hyperparameter set and the second hyperparameter set
Strubell teaches:
outputting a visual graph of the first energy consumption and the second energy consumption relative to the first hyperparameter set and the second hyperparameter set (Section 5, Table 3, “Our experiments suggest that it would be beneficial to directly compare different models to perform a cost-benefit (accuracy) analysis. To address this, when proposing a model that is meant to be re-trained for downstream use, such as retraining on a new domain or fine-tuning on a new task, authors should report training time and computational resources required, as well as model sensitivity to hyperparameters. This will enable direct comparison across models, allowing subsequent consumers of these models to accurately assess whether the required computational resources,” i.e., wherein a visual comparison show energy for different model configurations based on hyperparameter settings (hardware).)
PNG
media_image1.png
333
1174
media_image1.png
Greyscale
Strubell and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Strubell it would have been obvious for a person of ordinary skill in the art to apply the teachings of Strubell to Yang before the effective filing date of the claimed invention to apply hyperparameter configurations in order to improve the efficiency of energy consumption for training a machine learning model (Strubell, Introduction, “As a result, training a state-of-the-art model now requires substantial computational resources which demand considerable energy, along with the associated financial and environmental costs. Research and development of new models multiplies these costs by thousands of times by requiring retraining to experiment with model architectures and hyperparameters.”)
Claim(s) 6-8 and 15-17, and 20 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz, Justus, Desislavov, and Strubell further in view of Stamoulis et al., (“HyperPower: Power- and Memory-Constrained Hyper-Parameter Optimization for Neural Networks.”)
Regarding claim 6 and analogous claim 15:
Yang, as modified by Schwartz, Justus, Desislavov, and Strubell teaches the method of claim 5.
Yang, as modified by Schwartz, Justus, Desislavov, and Strubell does not explicitly teach:
wherein the visual graph further includes variations of the first energy consumption and the second energy consumption relative to quantities of the one or more epochs
Stamoulis teaches:
wherein the visual graph further includes variations of the first energy consumption and the second energy consumption relative to quantities of the one or more epochs (Section 3.1, Figure 3, “Figure 3. Visualizing our insights: how power varies vs accuracy with the number of training epochs (left); how accuracy can indicate configurations that do not converge to high-accuracy values (> 10%) (right).”)
PNG
media_image2.png
171
174
media_image2.png
Greyscale
Stamoulis and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Stamoulis it would have been obvious for a person of ordinary skill in the art to apply the teachings of Stamoulis to Yang before the effective filing date of the claimed invention to apply hyperparameter configurations in order to improve the efficiency of energy consumption for training a machine learning model (Stamoulis, Abstract, “While selecting the hyper-parameters of Neural Networks (NNs) has been so far treated as an art, the emergence of more complex, deeper architectures poses increasingly more challenges to designers and Machine Learning (ML) practitioners, especially when power and memory constraints need to be considered.”)
Regarding claim 7 and analogous claim 16, and 20:
Yang, as modified by Schwartz and Justus, teaches the method of claim 1.
Yang and Justus does not explicitly teach:
estimating, by the device, a plurality of accuracy values associated with corresponding quantities of epochs, for the machine learning model, based on the first hyperparameter set; and
determining, by the device, a plurality of energy consumptions, including the first energy consumption, associated with training the machine learning model and corresponding to the plurality of accuracy values
Stamoulis further teaches:
estimating, by the device, a plurality of accuracy values associated with corresponding quantities of epochs, for the machine learning model, based on the first hyperparameter set; and (Section 3.1, Figure 3, “Figure 3. Visualizing our insights: how power varies vs accuracy with the number of training epochs (left); how accuracy can indicate configurations that do not converge to high-accuracy values (> 10%) (right).”)
PNG
media_image2.png
171
174
media_image2.png
Greyscale
Schwartz further teaches:
determining, by the device, a plurality of energy consumptions, including the first energy consumption, associated with training the machine learning model and corresponding to the plurality of accuracy values (Page 54, key insights, “An alternative is Green AI, which treats efficiency as a primary evaluation criterion alonside accuracy. To measure efficiency, we suggest reporting the number of floating-point operations required to generate a result.”)
The motivation for claim 7 is the same as the motivation for claim 6.
Regarding claim 8 and analogous claim 17:
Yang, as modified by Schwartz, Justus, and Stamoulis teaches the method of claim 7.
Yang, as modified by Schwartz and Justus, does not explicitly teach:
outputting a visual graph of the plurality of accuracy values relative to the plurality of energy consumptions
Stamoulis further teaches:
outputting a visual graph of the plurality of accuracy values relative to the plurality of energy consumptions (Section 3.1, Figure 3, “Figure 3. Visualizing our insights: how power varies vs accuracy with the number of training epochs (left); how accuracy can indicate configurations that do not converge to high-accuracy values (> 10%) (right).”)
PNG
media_image2.png
171
174
media_image2.png
Greyscale
The motivation for claim 8 is the same as the motivation for claim 6.
Claim(s) 9 and 18 is/are rejected under 35 U.S.C 103 as being unpatentable over Yang, as modified by Schwartz, further in view of Hestness et al., (“Beyond Human-Level Accuracy: Computational Challenges in Deep Learning.”)
Regarding claim 9 and analogous claim 18:
Yang, as modified by Schwartz, Justus, and Stamoulis teaches the method of claim 8.
Yang, as modified by Schwartz, Justus, and Stamoulis does not explicitly teach:
indicating, on the visual graph, a portion associated with an inflection point
Hestness teaches:
indicating, on the visual graph, a portion associated with an inflection point (Page 8, Section 5.2 and figure 11, “We assume maximum achievable throughput of 80% of peak FLOPs and 70% of peak memory bandwidth, consistent with existing hardware. The accelerator’s compute intensity inflection point between memory-bound and bandwidth-bound (its Roofline “ridge point”) is 17.4 FLOP/B, but given peak achievable throughput, rises to 19.9 FLOP/B. We start by assuming that the accelerator has infinite memory capacity and is able to fit the memory footprint for a training step of any model.”)
PNG
media_image3.png
350
495
media_image3.png
Greyscale
Hestness and Yang are both related to the same field of endeavor (i.e., energy consumption). In view of the teachings of Hestness it would have been obvious for a person of ordinary skill in the art to apply the teachings of Hestness to Yang before the effective filing date of the claimed invention in order to improve the efficiency of energy consumption for training a machine learning model (Hestness, Abstract, “We further characterize and project the computational requirements to train these applications at scale. Our characterization reveals an important segmentation of DL training challenges for recurrent neural networks (RNNs) that contrasts with prior studies of deep convolutional networks. RNNs will have comparatively moderate operational intensities and very large memory footprint requirements.”)
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. THIS ACTION IS MADE FINAL. 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 AMINA BENOURAIDA whose telephone number is (571)272-4340. The examiner can normally be reached Monday-Friday 8:30am-5pm ET..
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, Michael J. Huntley can be reached at (303) 297-4307. 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.
/AMINA MORENO BENOURAIDA/Examiner, Art Unit 2129
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129
Prosecution Insights