Senior Software Engineer - AI & Data Platform

About the role

This is fundamentally a senior software/data systems engineering role. You’ll build the analytical substrate that AI agents operate over, then help design how those agents use it.

Trace is three layers, and this role works across all of them:

The metrics modeling layer. The engine that turns raw warehouse data into a queryable, mathematically consistent metric tree — the deterministic foundation everything else depends on.

The analysis layer. Driver decomposition, mix effects, anomaly detection, segment attribution — the computed primitives that produce an answer, not a plausible-sounding one. This is where most of the hard math lives.

The MCP server and agent harness. The tool surface and contracts that let agents query and reason over customer trees, the workflows behind our recurring briefings, and the evaluation harness that keeps their output correct at customer scale.

About You

To thrive in this role, you have

Extensive experience building and scaling data-intensive applications — pipelines, query engines, or performant data products.

Proven track record within a data product startup or as an engineer within a platform team.

Strong quantitative intuition and comfort reasoning about metrics, decompositions, ratios, and attribution. You care that an analysis ties out and can quickly learn the algorithms behind it.

Experience building production LLM or agent systems, ideally where agents operate over structured data and correctness matters.

High proficiency in TypeScript and Python — our metrics computation and query generation span both, over DuckDB and Postgres.

Strong SQL knowledge, and an instinct for where correctness breaks down at scale.

Drive for understanding and improving every aspect of the product, from infra to the analysis a stakeholder actually reads on Monday morning.