Research Scientist — Protein Engineering & Genome Editing

Protein engineering and genome editing sit in an awkward place for hiring. The work is unmistakably skilled, but the skill lives in the hands and in the build log, and almost none of it is legible from a job title. "Postdoctoral Researcher" covers the person who ran a screen someone else designed and the person who architected a switchable Cas construct, evolved it through four rounds, and worked out why rounds two and three failed. Those are not the same hire, and no title search separates them.

This page is for the second kind of researcher: PhDs, postdocs and research staff who design and build. If you can point at a construct and say what you chose, what broke, and what you changed, this is written for you.

What the roles actually look like

The companies hiring here are mostly venture-backed therapeutics platforms — precision oncology, cell and gene therapy, programmable biologics — at the stage where the platform is real but not yet finished. They are not looking for someone to supervise a pipeline. They want a scientist who owns the design-build-test cycle: designing engineered constructs, cloning them, iterating on specificity and control, and generating the material that decides which programme advances.

What they consistently struggle to find is not knowledge of CRISPR. It is demonstrated ownership of a fast, well-documented build cadence — someone who has personally driven many iterations and can explain the failures. That is a hard thing to advertise for, because every applicant's CV says "CRISPR".

Why a publication record works better here than keywords

Here is the specific failure this domain produces, over and over.

A scientist spends four years running directed evolution on an enzyme for industrial biocatalysis. They design variant libraries, build them with Golden Gate, screen thousands of candidates, and get a twenty-fold specificity improvement. Every capability a Cas-construct engineering role needs, they have done at depth.

They will not appear in any search for gene-editing candidates. Their papers are in different journals, they attend different conferences, and their CV contains none of the words a recruiter is searching for. The two fields share almost no vocabulary and almost all of the underlying craft.

Reading the publication record instead of the title finds that person. Not because it matches "directed evolution" as a string, but because the work described demonstrates iterative variant generation and specificity optimisation, which is what the role needs. This is the single clearest case in deep-tech hiring where keyword search structurally cannot do the job, and it is why domain transfer is worth building for rather than treating as an edge case.

What registering does, and does not, do

You give us an ORCID iD or a Google Scholar profile. We read the publication record and build a capability profile from what the work demonstrates. When a client role fits it, we get in touch.

There is no application to write and no need to be looking. You may hear nothing for months, because this depends on a matching role opening rather than on anything you do. Nothing about you is published, and we contact no one on your behalf without asking you first.

What we look for

• Modular molecular cloning — Gibson Assembly, Golden Gate and equivalents
• CRISPR/Cas construct design, beyond use of commercial systems
• Directed evolution and systematic variant generation
• Conditional, switchable and inducible protein systems
• Base-editing and prime-editing construct design
• Guide library design and off-target profiling
• Mammalian and bacterial expression systems, stable cell lines

Roles you might be matched to

• Research Scientist building Cas-based therapeutic constructs at a precision-oncology company
• Protein Engineer on a directed-evolution platform team
• Molecular Biologist owning the design-build-test cycle at a gene-editing startup

How this works

Register once with your ORCID iD or Google Scholar profile. We read your publication record and build a capability profile from what your papers show you can actually do. When a client role fits that profile, we get in touch.

This is not a single vacancy. You are registering for Protein Engineering & Genome Editing roles as they open, and there may be no suitable role today. Nothing about you is published, and we do not contact anyone on your behalf without asking first.

Questions

My directed evolution work was on industrial enzymes, not therapeutics. Does that transfer?

Often yes, and it is the case this approach exists for. Iterative variant generation, specificity optimisation and a fast build-test cadence are the same capabilities whether the target is a biocatalyst or a therapeutic construct. A keyword search for CRISPR therapeutics will never return you; reading what your papers show you actually did will.

I am hands-on wet lab rather than computational. Is this for me?

Yes. A lot of these roles specifically want the person who builds and troubleshoots rather than the person who models — someone who can say what they designed, what failed, and why the next iteration worked. That ownership shows up in a publication record.

Do I need to be actively job hunting?

No. Most people who register are not applying anywhere. Registering means we contact you when a role fits, not that you are on the market.

What happens to my ORCID iD?

We use it to pull your publication record and build a capability profile. It is how we match on research depth rather than on job titles.

Will my current employer see this?

No. Registrations are private. Nothing about you is published, and we do not contact anyone without asking you first.