What the work actually involves
You will spend most of your time doing two things: authoring molecular biology problems hard enough that a frontier model gets them wrong, and writing the fully documented correct solution alongside them. Concretely, that means specifying primer sets for PCR, qPCR and cloning with stated Tm, GC content and specificity checks; laying out expression vectors and the cloning strategy behind them (Gibson, Golden Gate, Gateway, classical restriction); selecting and scoring gRNA/sgRNA for CRISPR edits including off-target reasoning; designing mRNA and RNA constructs with UTRs, Kozak sequences and codon optimization; and building HDR or repair templates with justified homology arm lengths.
The other half is evaluation and standards work. You review model outputs against expert-quality answers and write feedback that explains why a design fails — a guide sitting in the wrong exon, a homology arm too short for the edit distance, a primer pair with a 3' complementarity problem. You also help write the rubrics other reviewers use, which means turning intuitions you normally apply silently at the bench into criteria someone else can apply consistently.
What the screen looks for
The platform screen is AI-led and probes depth by follow-up. Expect to be asked for specific constructs you personally designed, the parameters you chose, and what went wrong when they failed. Publication record matters here and is checked — first-author work in Nature-family journals, Cell, eLife, NAR, PNAS or EMBO carries the most weight, so include links. Prior AI training, red-teaming or annotation experience is a plus, not a gate. Candidates who describe running protocols someone else designed tend not to advance; the role is specifically about pre-experiment design decisions.
Logistics
- Remote, W-2 employment with Cincinnatus LLC (or its international entity), placed within a client AI lab's extended workforce
- Minimum 20 hours per week, on weekdays — this is a stated requirement, not a preference
- Observed pay band $70–105/hr; placement and rate depend on the screen and the client team
- Collaborative: you work alongside other SMEs, so consistency with shared guidelines matters as much as individual correctness