Archive Intel captures, archives and supervises business communications for regulated financial firms. Advisors and their compliance teams rely on us to preserve every client conversation, produce it on request, and flag what needs review. We're a small, fast-growing engineering team, and every engineer owns a meaningful part of the product.
AI flagging decides which messages a compliance reviewer needs to see. It's the core of our product. Our enterprise clients and their auditors want proof that it works and that it's done securely. We're hiring a hands-on engineer to own AI quality and safety end to end. You'll report to the CTO and work closely with our product engineers.
What you'll do
- Build and maintain labeled test sets with our compliance specialists, measuring missed issues, false alerts, latency and cost.
- Own the release process for models and prompts, backed by those tests.
- Evaluate new models, and lead upgrades when the results justify them.
- Harden the system against known vulnerabilities, keeping the system safe and secure.
- Turn reviewer decisions into a feedback loop that improves detection over time.
- Document how the system is tested and controlled, and represent it in client reviews.
What you'll bring
- Production backend experience in Python or TypeScript.
- An LLM feature you shipped to real users, and the tests that told you whether it worked.
- Startup experience owning a system end to end, with little process and real customers depending on it.
- A time your tests caught a regression before release, and what you changed.
- Regulated data or model risk documentation experience is a plus.
Benefits
Benefits include Health, Dental, Vision, Insurance, PTO, and 401k.
Equal opportunity
Archive Intel is an equal opportunity employer. We make employment decisions based on qualifications and business needs, without regard to race, color, religion or creed, sex, pregnancy, lactation, sexual orientation, gender identity or expression, national origin, ancestry, citizenship or immigration status, age, disability, genetic information, marital or partnership status, familial status, caregiver status, military or veteran status, height, weight, arrest or conviction record, credit history, unemployment status, sexual and reproductive health decisions, status as a victim of domestic violence, stalking or sex offenses, or any other characteristic protected by federal, state or local law. We will consider qualified applicants with arrest and conviction records in line with the NYC Fair Chance Act. If you need a reasonable accommodation to apply, email careers@archiveintel.com.