AI Biologist - Pathogen Countermeasures

AI Biologist - Pathogen Countermeasures

Latch builds rigorous scientific benchmarks (benchmarks.bio) for AI agents across biology. We work with frontier labs and pharma to measure whether agents can handle real biological workflows: the messy, multi-layer reasoning that matters in the field.

We’re seeking an AI Biologist to join our Pathogen Countermeasures team, pushing what AI can achieve in vaccine, antibody, and antiviral development against respiratory viruses.

About the Role. You will be the scientific authority on ground truth, testing whether agents can reason like an expert countermeasures scientist. The benchmark covers vaccines, therapeutic antibodies, and antivirals for influenza, RSV, and coronaviruses. AI agents today struggle to interpret raw neutralization, epitope, and drug-response data, to separate real escape or resistance from assay noise, and to make the advance-or-kill calls an expert would stand behind. You’ll work with scientists and engineers to build the tasks and datasets behind our benchmarks, and your judgment will set the standard agent outputs are held to.

Build the task

• Turn real countermeasure workflows into tasks with clear, gradable answers: antigen and immunogen comparison, epitope mapping, neutralization and cross-reactivity, antiviral susceptibility, and resistance interpretation

• Write gold-standard answers from raw data, with tolerances that reflect real assay, replicate, and strain variability

• Curate data: neutralization titers, binding and epitope maps, serology time courses, EC50 and dose-response curves, and viral sequences with resistance markers

Set the hard calls

• Test whether agents can tell real escape or resistance from artifact: a titer drop that’s really assay drift, a passenger mutation mistaken for a resistance marker, or a single fold-change not backed by sequence data

• Test whether agents separate binding from neutralization, immunogenicity from protection, and lab resistance from clinical relevance

• Test whether agents account for immune imprinting, waning, and cross-resistance when ranking candidates

Judge the agent

• Challenge conclusions that sound right but wouldn’t survive a program review

• Pinpoint where agents go wrong, what an expert would do instead, and what evidence would change the call

• Reward explicit uncertainty over confident guesses

Requirements

• 3+ years of hands-on experience in either vaccine/antibody development or antiviral development, ideally in respiratory viruses (influenza, RSV, or coronaviruses)

• Deep expertise in at least one of the areas below, in priority order within each track: Vaccines and antibodies

  • Vaccine antigen design, immunogen engineering, or therapeutic antibody discovery

  • Epitope mapping, neutralization, cross-reactivity, and immune recognition

  • Comparative evaluation of vaccine or antibody candidates from experimental data

  • Antibody responses, immune imprinting, or waning immunity

  Antivirals

  • Antiviral susceptibility and resistance, including phenotypic drug-response and viral sequence data

  • Antiviral discovery, mechanism-of-action studies, or experimental evaluation of candidates

  • Treatment-emergent resistance, cross-resistance, and resistance-marker interpretation

  • Antiviral assay development, validation, and dose-response analysis

• PhD in immunology, virology, microbiology, or a related field, or equivalent industry experience

• You’ve analyzed the data yourself, not just supervised it

• AI-ready: daily use of Claude, ChatGPT, or similar, with the judgment to spot answers that are technically correct but scientifically wrong

• Confidence to overrule an AI agent, and openness to update when shown wrong

Culture @ Latch

We practice radical transparency: you’ll know how the company is doing, what’s working, and what isn’t. It’s always us against the problem, never each other. We’re opinionated but respectful of each other’s time, and we think through possible solutions before bringing a problem to anyone. We use AI constantly, so we reward original expert judgment, not AI output passed off as your own. You’ll start on the bench that fits your expertise and can move across teams as you build a track record.

Outside of work: company outings most months, team off-sites, and a community that does cycling, soccer, figure skating, boxing, running, martial arts, music, books, and game nights.

Compensation & Logistics

Pay: $140K–$180K OTE (on-target earnings), with no ceiling.

• Your pay is your OTE times your weekly quota attainment. Hit quota, earn your OTE; double it, earn double. Example: $150K OTE at double quota = $300K.

• Your first 4 weeks (ramp) are paid at full OTE while we train you.

• You’re paid full OTE whenever quota can’t fairly apply: when a new project is starting, the platform goes down, or we pull you onto a special project.

In the US: full-time W-2

• Guaranteed base set by your state, paid every paycheck no matter what

• Health, dental, vision: 100% covered for you, 50% for dependents

• 401(k) (no match)

• PTO: 12 days to start, more as you’re promoted

Outside the US: contractor

• Paid on performance, with a monthly advance of 60% of OTE that’s never clawed back. US benefits don’t apply.

Growth: two tracks, based on output, not tenure.

• Individual contributor: six levels, so up to five promotions, with OTE rising at each one.

• Management: after your first promotion, you can move into managing a pod if that’s what you want.

Location: fully remote is fine; onsite is preferred, at our Mission Bay waterfront office in SF with catered lunch and dinner and unlimited snacks.

Work authorization: we sponsor H-1B (new and transfers), O-1, and TN, and support STEM OPT.

Interview Process

1. Apply. If your profile matches, we’ll send you a take-home project.

2. Take-home project (HackerRank). We review it as soon as it arrives.

3. Technical interview with a Member of Technical Staff (pod lead), scheduled right after your project review.

4. Culture interview with our C-suite.

5. Offer.

6. Background check, once you accept.