I build LLM-powered agent systems — agents that reason over real company data and actually answer the question you asked.
Right now I'm an Agent Engineer at Sharksia, working on production agents with LangChain and LangGraph on Azure. Before that I spent six years across fullstack web development and C++ systems programming, which turns out to be useful: most of the hard parts of an agent are the data and the plumbing, not the prompt.
- 🤖 RAG pipelines, retrieval strategies, and agent orchestration
- 🗄️ Comfortable in the database layer — SQL, PostgreSQL, pgvector
- 🔍 Big on observability: if you can't trace it, you can't improve it
- 📚 Currently going deeper on evaluation and retrieval quality
🤖 ceshar-ia — ask my CV anything
A conversational RAG agent that answers questions about my background, grounded in my actual CV and project documentation. Built with Python, FastAPI, LangChain/LangGraph, including a custom MCP server for tool integration.
Deployed on Railway with two surfaces: a REST API powering my portfolio, and a Discord bot.
→ Try it live at ccharris.vercel.app
My personal site — Astro + Sanity headless CMS, deployed on Vercel. Built for speed and easy content management, with ceshar-ia embedded so visitors can explore my work by asking instead of scrolling.
Structured Reports for DICOM medical images, in C++/Qt. From my time building PACS software at the Medical Biophysics Center.
AI / Agents
Data
Cloud & Tools
Web
📫 Open to conversations — LinkedIn · ccharris94@gmail.com