Job Description

The company

We are a stealth stage startup building data infrastructure for a large, underserved financial vertical. We take public records that today are slow, fragmented, and hard to trust, and turn them into a fast, reliable decision product that professionals will pay for. The company is a Delaware C corp, funded by founder capital, raising a pre seed round in 2026. The full thesis, market, and product are shared under NDA once we are in conversation.

The role

You are the support of the first engineer. You, the CTO are the entire engineering team through pre seed. This is not a narrow role. You own systems end to end: data ingestion, the data model, the ML that powers the core product, and the product surface a customer touches.

If you want a defined box, this is the wrong job. If you want to build a real system from the first commit and watch customers use it within months, this is the right one.

What you will do

• Build and operate the data pipelines that pull large volumes of public data into a clean, queryable store.
• Design the data model that turns raw records into trustworthy entities other systems depend on.
• Build the ML layer that links and scores records at scale, with measurable accuracy.
• Ship the serving API and the first product surface customers use.
• Set the engineering patterns the company builds on for years.

Work runs in two week cycles. Something real ships every cycle.

What we look for

• 1-2 plus years building production systems, with real depth in data engineering, ML, or both.
• You have built and run data pipelines other people depend on.
• You can carry a problem from messy source data to a working product feature with no one architecting it for you.
• You write code that holds up, and you make sound calls under ambiguity.
• You want the responsibility of being early, not the comfort of a defined box.

Strong pluses

• Record linkage, entity resolution, or graph data experience.
• Background in fintech, credit, real estate, or public records data.
• You have been an early engineer or founder at a startup before.

Stack

• Data: Python, dlt for ingestion, SQLMesh for transforms, DuckDB and DuckLake on S3 as the lakehouse.
• Orchestration: Dagster.
• ML: Splink and Python for record linkage and scoring.
• Backend: FastAPI, Postgres, Clerk for auth.
• Frontend: Next.js and TypeScript.
• Infra: Pulumi for infrastructure as code, AWS.
• Tooling: Codex and Claude Code used heavily; a small team shipping like a large one is the point.

You do not need every tool on this list. You do need to be the kind of engineer who picks one up fast.

Compensation

This is an unpaid intern role - stock based compensation is something that will be considered based on revenue hurdles following the public launch of the product.

Logistics

New York City, in person, full time. Reports to the CTO.

How to apply

Send a short note and one thing that shows how you think and build: a project, a system you are proud of, or code. We share the full company thesis under NDA early in the process.

Job Tags

Similar Jobs