30 years moving fuel across continents built the systems thinking. Now I build the AI infrastructure that makes complex operations intelligent.
60K+ lines of code across 15 repositories. Open to opportunities.
Multi-agent orchestration, LLM infrastructure, and production AI systems
Modular monorepo for production AI infrastructure. LLM gateway with multi-provider routing, semantic caching, and circuit breakers. HNSW vector search built from the original paper. ReAct agent runtime with tool registry. Document ingestion pipeline. Full observability with cost tracking and alerting.
Modular AI assistant framework with 12 independent modules. 5-tier LLM routing (Claude, Gemini, Ollama) with automatic fallback, persistent memory with hybrid search, animated avatar, desktop GUI, and extensible tool framework.
AI-powered customer operations platform. Chatbot with knowledge search, support rep copilot with AI-suggested responses, customer health scoring with churn prediction, ticket auto-categorization, onboarding flows, and sentiment analytics.
Automated compliance management for SOC 2, GDPR, and HIPAA. Gap analysis against 37 real regulatory requirements, document scanning, AI policy generation, drift monitoring, alerting, and audit-ready report generation.
Ingest company documents, TF-IDF vector search, AI-powered Q&A with source citations, auto-generated training articles and quizzes, and knowledge gap analytics dashboard.
Multi-tenant fleet management platform with role-based portals, real-time dispatch, franchise management, market simulation engine, and financial analytics. Monorepo architecture with shared design system.
Suite of developer tools: promptlab (pytest for LLM prompts), ragtest (RAG evaluation suite), fineforge (QLoRA fine-tuning pipeline for consumer GPUs), mcplex (MCP server for local AI models). All pip-installable with CLI interfaces.
Technologies I work with daily
I design and build the orchestration layer between raw LLMs and useful products — agent runtimes, LLM gateways, voice pipelines, and the infrastructure that makes AI systems reliable in production.
30 years and 3.5M+ miles in transportation operations built the systems thinking. 60K+ lines of code across 15 repositories turned it into software. I work across the full stack — Python, React, TypeScript, FastAPI — with deep focus on multi-agent coordination, LLM routing, and real-time pipelines.
Currently seeking roles in AI orchestration, agent infrastructure, and production AI systems.
Let's connect