Prosecution Insights
Prosecution Insights
Tech Center 4100 • Art Units: 2123 2124 2127 2129 2148 4100
This examiner grants 63% of resolved cases
Robert H. Bejcek II is an examiner in Art Unit 2123 of Tech Center 4100, with a total of 258 resolved cases. The allowance rate pct of 62.8 indicates a favorable likelihood of success for applicants, as more than three-fifths of cases reach a grant. This 62.8 percent rate suggests a positive environment for securing patent rights. However, the median days to allowance is 1393, which reflects an exceptionally long prosecution timeline for successful cases. This 1393-day median indicates that practitioners should be prepared for an extended period of examination that typically lasts much longer than the average. The 258 resolved cases provide a solid basis for these performance metrics.
The interview lift pct of 22.5 is a key metric for developing a successful strategy. This substantial lift indicates that engaging in an interview is highly correlated with an increased probability of allowance compared to cases without verbal negotiation. By utilizing an interview, applicants can significantly improve their chances of success beyond the 62.8 percent baseline. Given the 258 resolved cases, this 22.5 percent lift is a reliable indicator that direct communication is an effective tool in Art Unit 2123. Practitioners should consider an interview as a primary tactic to overcome obstacles and potentially expedite the 1393-day median wait time. The data underscores the importance of proactive advocacy in this examination environment. Effectively leveraging the interview process can help move applications toward a favorable outcome more efficiently.
| App # | Title | Status | Assignee |
|---|---|---|---|
| 17857731 | METHOD AND SYSTEM FOR ON-DEVICE INFERENCE IN A DEEP NEURAL NETWORK (DNN) | Final Rejection | SAMSUNG ELECTRONICS CO., LTD. |
| 17405875 | SYSTEM AND METHOD FOR MACHINE LEARNING BASED GROUND TRUTH LABEL EVALUATION IN TRAINING DATA SETS | Non-Final OA | Samsung Electronics Co., Ltd. |
| 18336589 | EFFICIENT PROCESSING OF TRANSFORMER BASED MODELS | Non-Final OA | Groq, Inc. |
| 18481685 | MULTI-MODEL INFERENCE PIPELINE AND SYSTEM | Non-Final OA | THE TORONTO-DOMINION BANK |
| 18702362 | VISUALIZATION METHOD, VISUALIZATION DEVICE, AND RECORDING MEDIUM | Non-Final OA | NEC Corporation |
| 17800305 | TRAINING DATA GENERATION DEVICE, TRAINING DATA GENERATION METHOD, AND PROGRAM RECORDING MEDIUM | Final Rejection | NEC Corporation |
| 17821076 | RESOURCE CONSTRAINED NEURAL NETWORK ARCHITECTURE SEARCH | Final Rejection | Google LLC |
| 16928573 | MACHINE LEARNING METHOD FOR THE DENOISING OF ULTRASOUND SCANS OF COMPOSITE SLABS AND PIPES | Non-Final OA | Saudi Arabian Oil Company |
| 18559102 | MONITORING AND/OR CONTROLLING A PLANT VIA A MACHINE-LEARNING REGRESSOR | Non-Final OA | BASF SE |
| 17458081 | TECHNIQUES FOR SCHEMA DRIFT DETECTION | Non-Final OA | Oracle International Corporation |
| 17828663 | ESTIMATING OPTIMAL TRAINING DATA SET SIZE FOR MACHINE LEARNING MODEL SYSTEMS AND APPLICATIONS | Final Rejection | NVIDIA Corporation |
| 16942507 | NEUROMORPHIC MEMORY AND INFERENCE ENGINE STACKED WITH IMAGE SENSOR TO REDUCE DATA TRAFFIC TO HOST | Non-Final OA | Micron Technology, Inc. |
| 18731339 | ADVANCED MODEL MANAGEMENT PLATFORM FOR OPTIMIZING AND SECURING AI SYSTEMS INCLUDING LARGE LANGUAGE MODELS | Non-Final OA | QOMPLX LLC |
| 18081515 | SYSTEM AND METHOD FOR MATHEMATICAL MODELING OF HARDWARE QUANTIZATION PROCESS | Final Rejection | Black Sesame Technologies Inc. |
IP Author analyzes examiner patterns and generates tailored response strategies with the highest chance of allowance.
Build Your Strategy