AI Claude Desk Developer

Hedge Fund based in Greenwich CT & NYC is looking for a desk-level developer to build the AI tooling layer that sits between investment professionals and their data. You will sit with the deal teams, watch how they actually work, and turn recurring analytical work — capital structures, pricing, exposure screens, issuer diligence, sector monitors — into durable, reusable tools that anyone on the desk can run.

What you'll do

You will design and ship Claude Skills and MCP servers that connect their models to Databricks, DealCloud, Box, Microsoft 365, and licensed market and credit data vendors. That means owning the full path: understanding the investment question, writing the SQL against our Unity Catalog gold layer, wrapping it in tooling that behaves reliably under real-world inputs, and iterating with the analyst who asked for it.

Who we're looking for

You should already be building with Claude. Hands-on depth in Claude Skills, MCP servers, and agentic tooling patterns is the core requirement — we care more about what you have actually shipped with these tools than about years of general software experience. Python, SQL, and Databricks are the substrate. Roughly 3-6 years building production data or analytical tooling is the right range.

You should be passionate about AI — following where the tooling is going, testing new releases as they land, and forming your own view on what is genuinely useful versus what is noise.

This role rewards someone who would be doing a version of it in their own time anyway.

Familiarity with credit or private equity markets — capital structures, spreads, leverage and coverage, fund-level exposure — is strongly preferred, as is prior experience supporting a trading, research, or investment desk. If you don't have it, you should be able to show that you learn a domain fast and enjoy doing so; the tools only work if you understand the question behind them.