Summary
Gregory Kennedy is a Lead AI Systems Architect based in Los Angeles who designs deterministic, regulatory-compliant AI/LLM systems—most notably Keraunos, a 10-stage deterministic agent pipeline with 64 custom tools and enforced human-in-the-loop escalation to ensure auditable, policy-perfect outputs. With eight years focused on production ML engineering and a Silicon Valley systems background, he distills large frontier models into efficient 3B–4B production students using a multi-tier teacher-student distillation and context-injection methodology. He pairs deep engineering (Rust, Python, PostgreSQL/pgvector, self-hosted VPS deployments) with applied research—authoring SYRP-DGM and a reputation-based ML algorithm—and has practical humanitarian ML experience from a Kaggle/DeepMind Gemma 3N hackathon and multilingual on-device apps. Gregory also trains professionals and develops courses for platforms like Pluralsight, blending enterprise-grade AI delivery with accessible education and a track record of fundraising and social-impact projects. An uncommon strength is his insistence on deterministic closure—automatic linting, security gates, and a two-attempt fix limit that forces human escalation—so clients get repeatable, auditable AI behavior rather than hopeful outputs.
8 years of coding experience
19 years of employment as a software developer
Bachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at DePaul University
German