Presentation
Deep work on the AI side: custom agent harness over Anthropic, OpenAI, Gemini, OpenRouter with tool use, function calling, structured outputs, streaming, dynamic model selection per task. Hybrid RAG with pgvector semantic search + keyword + AI reranking + citation grounding. Custom StreamJsonTokenizer that reconstructs streaming JSON token-by-token for real-time structured UI population. Internal eval suite + A/B harness for prompt/model comparison.
Not a research scientist. I'm an AI systems engineer. I build the layer between the model and the product, and I take it to production.
U of C CS grad with Distinction (3.97 GPA). Stack: TypeScript, Python, Next.js, AWS, pgvector.
