Molecular Biology Expert (US)

What you'll do

The qualification that matters is in the first paragraph of the ad: they want people who have designed DNA and RNA sequences themselves, before the experiment ran. Having chosen the homology arm length and lived with the result is a different job from reading about construct design, and the listing is written for the first.

The work is finding where a model's molecular biology reasoning breaks down and closing the gap. That means designing hard, domain-relevant tasks and writing the accurate, well documented solutions that become ground truth: primers for PCR, qPCR and cloning, plasmids and expression vectors, gRNA and sgRNA for CRISPR editing, mRNA and RNA constructs, and HDR or repair templates. You review tasks and model outputs against expert-quality solutions and give written feedback on correctness, rigour and biological reasoning, and you work alongside other subject matter experts so the training data stays consistent.

The part with the longest half-life is the rubric writing. You author the guidelines used to judge sequence design quality, covering guide and target selection, homology arm length, primer melting temperature and specificity, codon optimisation and regulatory elements. That is encoding what counts as a defensible design decision, and those criteria outlive any individual task.

What you need

• An advanced degree, PhD strongly preferred, in molecular biology, genetics, biochemistry, synthetic biology, bioengineering or a related life science field
• Hands-on experience designing nucleic acid constructs, across several of primers, plasmids and vectors, gRNA and sgRNA, mRNA, and HDR or repair templates, together with the cloning strategy behind them (Gibson, Golden Gate, Gateway, restriction) and codon optimisation
• A strong peer-reviewed publication record weighted toward first-author work. The ad names Nature and the Nature family, Cell, eLife, Nucleic Acids Research, PNAS and EMBO Journal as the kind of venue it means, and asks you to include links to your publications when you apply
• Clear career progression
• A reliable commitment of at least 20 hours a week, on weekdays
• Strong written communication, with the ability to explain the reasoning behind a design choice clearly and precisely

Previous work in AI training, model evaluation or data annotation is a plus rather than a requirement. Eligibility is the United States.

How you would be employed: read this twice

The ad describes two incompatible arrangements on one page, and which applies changes your tax, your benefits and what the hourly rate is actually worth.

The role section says this is a W-2 employment position with Cincinnatus LLC, an enterprise staffing company acting as employer of record and providing payroll, benefits and compliance, placing you inside a leading AI lab's extended workforce. The contract and payment terms at the bottom of the same listing say you will be engaged as an independent contractor, fully remote on your own schedule, paid weekly via Stripe or Wise.

W-2 employment means withheld tax and employer-provided benefits; contractor status means neither, and you carry self-employment tax on the same headline number. Get the answer in writing before you accept, because $105 an hour as a W-2 employee and $105 an hour as a contractor are not the same offer.

What it pays

$70–105 per hour, Mercor's own published figure rather than an estimate of ours, and high for life science work that does not require leaving your current position.

At the stated minimum of 20 hours a week, that band works out at roughly $1,400 to $2,100 a week. Treat that as arithmetic rather than a promise: the ad also states that projects can be extended, shortened or concluded early depending on need and performance, so the commitment runs one way, and the employment ambiguity above determines what lands in your account after tax.

The band is $35 wide with no published basis for placement. Publication record and the breadth of construct types you have designed are the plausible levers, but the ad does not say so.

Worth knowing

Good:

• $70–105/hour is strong for life science contract work, and the skill being paid for is one you already use weekly
• Authoring the rubrics is real intellectual work, and it is the part you can describe afterwards without breaching anything
• Fully remote, and the work maps directly onto what you do before an experiment runs
• Benefits, payroll and compliance are on the table if the W-2 route is the one that actually applies
• No prior AI or ML experience required
• Working alongside other SMEs rather than alone in a queue

Less good:

• The employment basis genuinely contradicts itself on the page: W-2 employment in one section, independent contractor in another
• A first-author record in a named list of high-impact journals is a pedigree filter that demonstrated bench skill does not get you past
• 20 hours a week, weekdays only, is a hard constraint alongside a postdoc or a lab position, and the weekday restriction removes the obvious workaround
• Projects can be extended, shortened or concluded early, so hours and duration are not guaranteed
• H-1B and STEM OPT candidates cannot be supported, which excludes a substantial share of exactly the US postdoc population this ad describes
• The AI lab is not named, so you cannot know whose model your ground truth trains
• Contractor tax and no notice period, if that is the arrangement that applies

About this listing

Posted by Mercor as an hourly remote contract restricted to the United States, confirmed open on 12 September 2026. The $70–105 band, the 20-hour weekly minimum, the journal list, the Cincinnatus LLC employer-of-record arrangement and every qualification above are the ad's own published text.

Mercor runs the same role for the United Kingdom at the same band. See Molecular Biology Expert (UK) if you are based there.

The W-2 and independent contractor statements both appear on the listing as published; we have not resolved the contradiction on Mercor's behalf, and no client is named. Mercor also states that the work will not involve access to confidential or proprietary information from any employer, client or institution, which is a constraint on you as much as a reassurance: unpublished constructs and data from your own lab are out of scope. Payment and onboarding run at the platform level. See Mercor.