Prosecution Insights
Prosecution Insights
Tech Center 2100 • Art Units: 2123 2125 2145
This examiner grants 74% of resolved cases
Steven Huynh Phung works in Tech Center 2100 and Art Unit 2123, where he has handled 46 resolved cases. The allowance rate is 73.9%, which is a relatively high figure. This 73.9% rate suggests a good probability of success for applicants. However, the median days to allowance is 1613, indicating a very slow prosecution process that lasts more than four years on average.
The interview lift of 30.2% is significant, suggesting that direct communication is a powerful tool for advancing applications. Given the long 1613-day median pendency, practitioners should use interviews to try and resolve issues more quickly. The 30.2% lift shows that interviews can substantially improve the already favorable 73.9% allowance rate. Applicants should be prepared for a lengthy timeline but can take confidence in the high likelihood of eventually reaching an allowance through persistent engagement and oral advocacy.
| App # | Title | Status | Assignee |
|---|---|---|---|
| 18485457 | Scalable Self-Supervised Graph Clustering | Non-Final OA | Google LLC |
| 17448298 | QUANTIZED FEEDBACK IN FEDERATED LEARNING WITH RANDOMIZATION | Non-Final OA | QUALCOMM Incorporated |
| 17592196 | SYSTEM AND METHOD FOR HETEROGENEOUS MULTI-TASK LEARNING WITH EXPERT DIVERSITY | Non-Final OA | ROYAL BANK OF CANADA |
| 17508734 | TECHNIQUES FOR TRAINED MODEL BIAS ASSESSMENT | Non-Final OA | Oracle International Corporation |
| 17729330 | MACHINE LEARNING BASED MONITORING FOCUS ENGINE | Final Rejection | Microsoft Technology Licensing, LLC |
| 17362684 | NEURAL NETWORKS WITH ANALOG AND DIGITAL MODULES | Non-Final OA | International Business Machines Corporation |
| 17519532 | NEURAL NETWORKS TRAINED USING EVENT OCCURRENCES | Final Rejection | NVIDIA Corporation |
| 18427666 | RE-ENGINEERING DATA TO ENABLE AI TO EXCEED ITS CURRENT LIMITS BY UTILIZING QUANTUM ENGINEERING | Non-Final OA | Bank of America Corporation |
| 18436313 | SYSTEMS AND METHODS FOR STRESS DETECTION USING KINEMATIC DATA | Non-Final OA | Board of Regents, The University of Texas System |
IP Author analyzes examiner patterns and generates tailored response strategies with the highest chance of allowance.
Build Your Strategy