Work on something with an answer key.

bearsignal.ai — /careers

Most software ships and you never find out if it mattered. Our output gets graded — by enforcement actions, by restatements, by what happens to a stock after we said the numbers don’t hold. We find out, publicly and on a delay we don’t control, whether we were right.

That’s the deal at BearSignal: a small team, an adversarial problem, and a scoreboard. We are building the infrastructure to run forensic mathematics against every listed company in the U.S. and Hong Kong — and the judgment system that learns from every outcome. If you’ve been looking for work where “did it work?” has an actual answer, keep reading.

Senior Engineer — Forensic Data Systems #

Silicon Valley (Bay Area) · onsite/hybrid

What you’ll build

You’ll own production systems where data engineering meets forensic finance: pipelines that turn two regulators’ worth of disclosure into mathematics-ready fundamentals; the scoring infrastructure that runs a growing family of detection models across ~10,000 listed companies every reporting period; and the data foundations of a judgment layer calibrated on real-world outcomes. You will not be maintaining someone else’s legacy decisions — the system is young, the architecture is deliberate, and your fingerprints will be on load-bearing walls.

What we’re looking for

• Deep PostgreSQL fluency — you think in schemas, you’ve moved data at COPY scale, and you have opinions about NULL
• Production Python; comfort with the scientific stack (numpy/pandas) and with code that implements published quantitative models faithfully
• Experience building data pipelines against messy real-world sources — regulatory filings, scraped documents, multi-language PDFs are all in scope
• AWS production experience (containers, managed Postgres, queue/batch patterns)
• Enough financial-statement literacy to read a balance sheet and care when it doesn’t articulate — or the demonstrated ability to learn domains fast
• Bonus: Mandarin reading ability; experience with LLM-based extraction at batch scale; a quant or forensic accounting background

How we work

AI does the heavy computation and the first draft of nearly everything; humans own judgment, taste, and the decisions that can’t be reversed. We write our rules down — data discipline, write boundaries, what the engine is allowed to conclude (nothing) — and the documents govern, not the loudest voice. Small team, founder-led, no meetings that should have been a schema comment.

Apply: careers@bearsignal.ai — send something you built and the decision you’re proudest of inside it. No cover letter needed.

Field Research Internship #

Tempe, AZ / Phoenix metro

We partner with universities to put student researchers in the field. The work: structured, narrowly-scoped evidence collection — photographs, site observations, public-record verification — executed against precise task specifications, in public settings, with safety protocols and full institutional backing. You don’t need a finance background; you need reliability, attention to detail, and the ability to follow a spec exactly. What you get: real fieldwork on real investigations, a window into how modern forensic research operates, and output that feeds directly into a production research system.

Currently onboarding through university partner programs. If your institution is interested: partnerships@bearsignal.ai.