ROY MAYNOR
Houston,TX
Summary
Principal Neural Architect with a focus on Fundamental AI Research (FAIR) and deep systems engineering. Expert in CUDA/Triton kernel optimization, agentic reasoning frameworks, and distributed high-performance computing. A rare career trajectory spanning three decades of technological evolution, from legacy systems management to state-of-the-art neural scaling and recursive self-correction loops.
Overview
28
28
years of professional experience
Work History
Senior ML Engineer
META AI (FAIR - Fundamental AI Research)
04.2020 - Current
- Throughput Optimization: Architecting core optimizations for low-level CUDA/Triton kernels within internal training libraries, enabling significant throughput gains for the largest-scale training runs (Llama-3/Llama-4).
- Agentic Frameworks: Leading a specialized research sub-team focused on 'Self-Correction' loops in autonomous agents, mitigating hallucinations in 128k+ long-context window tasks.
- Scale Engineering: Orchestrating cross-departmental data pipelines processing 5PB+ of unstructured data daily for foundational model pre-training.
Graduate Research Assistant (HCI & AI)
STANFORD UNIVERSITY
09.2015 - 03.2020
- Collaborated on early Transformer research and scaling law analysis.
- Developed 'Vessel' (v1.0), a bridge for local-to-cloud agentic execution.
Lead Software Engineer
INTEGRATED SYSTEMS CORP
01.2008 - 08.2015
- Transitioned the company from monolithic PHP/Perl architecture to a distributed Microservices model.
- Managed the first internal adoption of Big Data pipelines using Hadoop and early Spark clusters.
Junior to Mid-Level Software Engineer
DATASTREAM SOLUTIONS
06.2002 - 12.2007
- Developed and maintained enterprise-scale SQL databases and early web-based portal systems.
- Self-taught C++ and Python during this period to automate internal server provisioning.
Senior Field Technician
GEEK SQUAD / PC SOLUTIONS
08.2000 - 05.2002
- Performed component-level hardware repair and network troubleshooting for corporate clients.
- Mastered the fundamentals of the OSI model and BIOS-level hardware optimization.
Service Champion & POS Liaison
TACO BELL
06.1998 - 07.2000
- Managed front-of-house operations and Point-of-Sale (POS) system stability during high-volume shifts.
- Optimized drive-thru throughput metrics via process streamlining, achieving top-tier efficiency rankings in the regional district.
Education
Dual M.S. - Computer Science & Symbolic Systems
STANFORD UNIVERSITY
B.S. - Computer Science
UNIVERSITY OF TEXAS AT DALLAS
Skills
- AI/ML: PyTorch, Triton, JAX, HuggingFace, RLHF, RAG, Quantization (FP8/INT4)
- Systems: CUDA, C, Rust, Go, Kubernetes, Slurm, Docker, GCP/AWS
- Legacy/Foundational: Java, Perl, PHP, SQL, Linux Kernel optimization
Timeline
Senior ML Engineer
META AI (FAIR - Fundamental AI Research)
04.2020 - Current
Graduate Research Assistant (HCI & AI)
STANFORD UNIVERSITY
09.2015 - 03.2020
Lead Software Engineer
INTEGRATED SYSTEMS CORP
01.2008 - 08.2015
Junior to Mid-Level Software Engineer
DATASTREAM SOLUTIONS
06.2002 - 12.2007
Senior Field Technician
GEEK SQUAD / PC SOLUTIONS
08.2000 - 05.2002
Service Champion & POS Liaison
TACO BELL
06.1998 - 07.2000
Dual M.S. - Computer Science & Symbolic Systems
STANFORD UNIVERSITY
B.S. - Computer Science
UNIVERSITY OF TEXAS AT DALLAS