MedInsider
Coming soon.
Ahmed Taha, intrigued by the opportunity to apply technical skills to the legal domain, became a Patent Examiner at the USPTO specializing in computer graphics and machine learning technologies. With a foundation in software engineering from nearly five years at District Hut, combined with his education in cybersecurity, machine learning, and quantum computation from Johns Hopkins and Columbia, he brings deep technical expertise to patent examination. Currently pursuing his M.S. in Computer Science at Columbia while conducting research at the intersection of AI alignment and the medical field. His research interests include computer vision, deep learning, and AI alignment.
Publications
SpineFairBench: A Counterfactual Benchmark for Auditing Demographic Sensitivity in Spinal Radiology VLM Reports
Preprint, 2026
TL;DRPaired counterfactual benchmark that audits whether nine frozen vision-language models change their spinal-radiology reports when apparent age and sex are edited while target pathology is preserved — measurable recommendation drift in all nine, with management recommendations less stable than diagnostic-label overlap under the same demographic edit.
Software
ahmedtaha100/SpineFairBench
Counterfactual benchmark auditing demographic sensitivity in spinal radiology vision-language model reports.
ahmedtaha100/LLM-Counsel
Multi-model LLM router with query complexity classification, dissent detection, and semantic caching.
ahmedtaha100/MoR
PyTorch implementation of Mixture-of-Recursions: recursive transformers with dynamic per-token routing (NeurIPS 2025).
ahmedtaha100/RRT
JAX/Flax implementation of Relaxed Recursive Transformers with layer-wise LoRA (ICLR 2025).
jax-ml/jax #36521
Added complex-input support to jax.scipy.special.gamma using the Lanczos approximation and reflection formula in Google JAX.
Azure/azure-sdk-for-python #46137
Fixed an azure-data-tables bug (#46014) where unsupported TableClient.list_entities(...) kwargs leaked into the async transport and surfaced as an aiohttp TypeError; added early SDK-level validation for query_filter/parameters with clear guidance to use query_entities, plus sync/async regression tests.
Azure/azure-sdk-for-python #46546
Submitted a follow-up fix for the broader transport-layer kwargs leak tracked in #46365, adding built-in transport validation across azure-core and corehttp so unsupported pipeline kwargs fail clearly before reaching aiohttp, requests, or httpx, while preserving custom transport passthrough.
Contact
Education
Work Experience
August 2025 - May 2027
M.S. Computer Science (Machine Learning)
Columbia University
Fu Foundation School of Engineering. Advanced coursework in machine learning, artificial intelligence, and computational systems.
January 2025 - Present
Patent Examiner
U.S. Patent and Trademark Office
Promoted from GS-9 to GS-11 January 2026. Specializing in computer graphics and machine learning technologies. Received commendation letter from OPQA director for quality of office actions. Achieved 104% production average.
August 2024 - August 2025
M.S. Cybersecurity
Johns Hopkins University
Whiting School of Engineering. 3.85 GPA. Coursework in Quantum Computation, Ethical Hacking, Web Security, and Cryptology. Published ML cryptanalysis research paper.
August 2024 - August 2025
Research Assistant
Johns Hopkins University – CCVL Lab
Whiting School of Engineering. Conducted research at the Computational Cognition, Vision, and Learning (CCVL) lab.
June 2020 - January 2025
Founder & Software Engineer
District Hut LLC
Led full-stack development projects generating over seven figures in revenue. Built patient-management systems, auto-dealer platforms, and restaurant applications. Registered trademark for company slogan.
2018 - 2024
B.S. Computer Science
California State University, Sacramento
3.45 GPA. Division 2 Wrestling Team. Overcame serious spine injury to complete degree. Developed Adapted Strength fitness platform as capstone project.
August 2017 - February 2021