AI Biologist - Cell Therapy Discovery

LatchBio•San Francisco, CA
•$140,000 - $180,000

About The Position

Latch builds scientific benchmarks for AI agents across biology and partners with frontier labs and pharma to evaluate agents on real biological workflows. The AI Biologist will join the Cell Therapy Discovery team and serve as the scientific authority on ground truth, assessing whether agents reason like expert cell therapy scientists. The role involves building benchmark tasks and datasets across CAR, TCR, and other engineered cell formats, spanning antigen selection through candidate nomination. The successful candidate will create gold-standard answers with defensible scientific rationales; design scenarios across the cell therapy discovery arc; curate datasets from real programs; and develop Python analysis pipelines where useful. They will identify experimental artifacts, distinguish activation from exhaustion, recognize tonic signaling, adjudicate conflicting results, and assess tradeoffs such as potency versus exhaustion and affinity versus off-tumor selectivity. They will evaluate agent reasoning step by step, identify errors and the evidence needed to change a conclusion, and score confidence while rewarding appropriate uncertainty. The role is available remotely or onsite at Latch's Mission Bay office in San Francisco, with onsite work preferred. US-based employment is full-time W-2; workers outside the US are engaged as contractors. The position offers compensation tied to weekly quota attainment, with a guaranteed ramp period and full OTE when quota cannot fairly apply.

Requirements

  • PhD in immunology, cell biology, bioengineering, or a related field; an MS with equivalent hands-on program experience also qualifies.
  • 2 to 6 years of hands-on cell therapy experience at a biotech or pharma, with personal analysis work in the last two years.
  • Personally generated or analyzed data used to decide which target or construct advanced, and able to name the threshold and decision.
  • Direct experience in at least one of the following: CAR or TCR construct design; multiparameter flow panel design and analysis; cytotoxicity and co-culture assays; cytokine profiling; or exhaustion and persistence characterization.
  • Deep expertise in one listed area and fluency across the others.
  • Strong experimental design, quantitative analysis, and statistics skills.
  • Daily use of Claude, Cursor, or similar AI tools, with the judgment to recognize scientifically unsound model reasoning and confidence to overrule it.

Nice To Haves

  • Python or R for personal analyses.
  • Experience with allogeneic or gene-edited cell therapy, NK or Treg formats, or in vivo CAR.

Responsibilities

  • Write gold-standard answers and defensible rationales for cell therapy benchmark tasks.
  • Design benchmark scenarios across antigen prioritization, receptor and construct design, phenotype, potency and killing, cytokine profiling, exhaustion and persistence, and on-target off-tumor triage.
  • Curate datasets from real programs, including flow and spectral panels, cytotoxicity readouts, cytokine multiplexes, repeat-stimulation and serial-killing assays, inhibitory receptor panels, memory phenotype distributions, and normal-tissue antigen expression.
  • Build Python analysis pipelines where useful, with tolerances that work across donors, runs, and readout platforms.
  • Identify experimental artifacts and distinguish genuine biological effects from confounds such as receptor expression differences, assay saturation, viability-readout interference, activation-induced cell death, control normalization, donor variation, and spillover errors.
  • Assess whether evidence supports claims about activation, exhaustion, tonic signaling, persistence, potency, and other cell therapy outcomes.
  • Adjudicate conflicting results and identify experiments that would resolve uncertainty.
  • Evaluate tradeoffs among potency, exhaustion, cytokine output, toxicity risk, affinity, selectivity, expansion, and memory phenotype.
  • Interrogate AI agent reasoning, identify technical or scientific errors, explain why conclusions fail, and specify what evidence would change the assessment.
  • Score agent confidence, rewarding explicit uncertainty when data are ambiguous.

Benefits

  • For US employees, health, dental, and vision coverage is 100% covered for the employee and 50% covered for dependents.
  • 401(k) with no match.
  • 15 days of PTO to start, with more as the employee is promoted.
  • First four weeks are paid at full OTE during ramp.
  • Full OTE is paid when quota cannot fairly apply, such as when a new project is starting, the platform is down, or the employee is assigned to a special project.
  • Company outings most months and team off-sites.
  • US employees receive a guaranteed floor (base salary) set by their state, paid on the 15th and last day of each month.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service