hey, i'm steven 👋
i build things with LLMs and try to make AI actually useful.
🔬 Socionics Research Lab — multi-agent AI system for personality typing. three specialist agents analyze from different angles, then an adversarial validator checks their work. built the prototype in python/streamlit, rebuilt it properly in next.js.
- multi-agent architectures and agentic workflows
-
- making complex systems understandable
-
-
personality typology (socionics specifically - 7+ years deep)
-
-
figuring out what AI is actually good for beyond demos
-
python javascript react next.js langchain streamlit
currently looking for roles where i can build things and solve problems — solutions engineering, dev rel, or similar. if you're working on something interesting, let's talk.
-
-