Erik Scott

TypeScript, Rust, Go, Python, React, Node, PostgreSQL. Vector DBs, graph databases, ML pipelines. 18 years full-stack, now building MCP servers and obsessing over latency.

LLM orchestration patterns, vector embeddings, and graph-based knowledge retrieval. Mostly solving how to make AI assistants actually remember and use context efficiently.

ContextStream - AI memory using vector search and knowledge graphs. Persistent context for AI so they use the most relevant memory and context for the user. Rust backend, TS MCP server.

Chats about AI/ML dev tools, full-stack architecture, or the indie hacker journey. Always down to talk shop about making developer workflows less painful.