Summary
Adrian Scott is an ML engineer and serial founder with 13 years of experience building production AI systems and startups from San Francisco. He combines a Ph.D. in nonlinear optimization with hands-on expertise in distributed PyTorch training, transformer fine-tuning, and automated trading systems, having driven 20x performance improvements and built 24/7 algorithmic hedge strategies. As founder or early leader at Ryze, CoderBuddy, Freedom Stack and Soma Capital, he has repeatedly bridged research, product and growth—once taking a music-tech startup from seed-stage investor relations (Napster early backer) to notable exits and traction. Currently he helps a small set of CEOs grow revenue by leveraging LLMs and bespoke ML tooling, and maintains private codebases focused on LLMs, STT, and multi-bot chat orchestration. His background uniquely mixes deep math, low-level performance tuning, and founder-level commercial instincts, making him comfortable shipping both high-performance models and market-facing products.
13 years of coding experience
32 years of employment as a software developer
Ph.D. Mathematics Nonlinear Optimization, Ph.D. Mathematics Nonlinear Optimization at Rensselaer Polytechnic Institute