AI Quality & Safety Engineer

Location: San Francisco (onsite)

Compensation: $120–170K + variable equity

About Lotus AI

Lotus AI is a primary care app that combines your medical records, AI, and real doctors to deliver free, personalized care (including prescriptions and referrals). We’re backed by Kleiner Perkins + CRV, and we’re a small team that ships fast.

The role

You will ensure Lotus feels safe, reliable, and works day-to-day for patients. This role sits close to product and engineering and blends quality, debugging, and building lightweight systems that catch issues before users do.

This is not traditional QA or customer support. It’s a hands-on, product-adjacent role focused on turning real user issues into fixes, better evaluations, and tighter systems.

What you’ll do

• Run evaluations and regression checks on real workflows before/after model or data changes.

• Track common failure modes and convert them into repeatable tests.

• Triage patient issues and product bugs, and help drive fast resolution.

• Partner with clinicians and engineers on safety/quality escalations.

• Build lightweight tooling and dashboards to surface what’s breaking.

What this role is not

• A scripted support queue.

• “Write test cases forever and never ship.”

• A role where everything is defined upfront.

This role is for you if

• You like debugging and figuring out what’s actually happening.

• You can write code (usually Python) and you’re comfortable with logs and messy data.

• You care about doing things carefully

• You enjoy being the glue between product, engineering, and users.

Bonus points

• Python and SQL.

• Comfort reading logs/traces, reproducing issues, and writing crisp bug reports.

• Familiarity with LLM evals, RAG, and prompting (or willingness to learn quickly).

• Strong judgment with sensitive data and HIPAA-aware habits.

Why Lotus

The work directly impacts real people. If you want to ship quickly and care about quality, this role is a strong mix.

Interview process

Quick intro call → short technical screen → deeper technical conversation → references → offer

We typically move in ~7–10 days.