As AI becomes more embedded in biotech R&D, companies are competing for professionals who can bridge biology, machine learning, data and engineering. Recruiting expert Darren Nelson explains which capabilities are becoming hardest to find—and why.
Biotech has always competed for highly specialized talent. A company working in gene therapy, oncology or rare disease cannot simply hire a strong scientist; it needs someone with deep expertise in the specific science behind its work. As artificial intelligence (AI) becomes more integrated into research and development, that talent challenge is becoming even more complex.
Some of the hardest people to find today are not traditional data scientists or traditional biologists. They are professionals who can work across both worlds: scientists who understand machine learning, engineers who understand biological data and technical leaders who can translate a scientific problem into something an AI team can actually build.
From where I sit in recruiting, that ability to connect the science and the technology is becoming more important than the title on someone’s resume.
The BioSpace 2026 U.S. Life Sciences Employment Outlook report found that 64% of organizations surveyed late last year were actively recruiting, with artificial intelligence among the areas employers were prioritizing. Benchling’s 2026 Biotech AI Report found a similar pattern. Among 100 biotech and biopharma organizations using AI, the top talent priority was data scientists with life sciences domain knowledge. Computational biologists with AI expertise followed, along with AI engineers who can build and deploy models. That makes sense, because the challenge in biotech is rarely just getting a model to work.
A software engineer can build an impressive machine-learning system and still miss something scientifically important about the underlying data. A bench scientist can understand the biology extremely well but may not know how to evaluate data leakage, model performance, infrastructure constraints or whether a system can be reproduced reliably. The most valuable people increasingly understand enough of both sides to recognize those problems before they become expensive.
Computational biology is a good example. The job is becoming more engineering-heavy. Companies still need people who can interpret biological data, but they increasingly also need scientists who can build durable pipelines, work with modern machine-learning frameworks, manage large datasets and collaborate with software and infrastructure teams.
At the same time, AI engineering in biotech is becoming more domain-specific. Experience building recommendation systems, consumer applications or generic large language model products can certainly transfer, but biological data presents its own set of problems. It can be noisy, sparse, multimodal and expensive to generate. Ground truth is not always obvious. In many environments, reproducibility and data provenance matter just as much as model accuracy. That changes what “good AI talent” means in a biotech company.
The same thing is happening with laboratory automation. As AI becomes more involved in experiment design and closed-loop discovery, the line between computational and experimental work gets less clear. The valuable person may be someone who understands the experiment well enough to know what data should be generated next, while also understanding how that data will affect the model.
We are also beginning to see early forms of agentic AI enter biomedical research. A 2026 Nature Biotechnology article described AI agents performing tasks ranging from literature review and hypothesis generation to data analysis and model interpretation. If that direction continues, biotech companies will need more people who can oversee workflows that combine software, scientific judgment and experimental decision-making.
There is another skill that I think gets overlooked: translation. AI projects often struggle because the scientists, engineers, data teams and executives are not actually describing the same problem. Someone has to be able to take a scientific question and turn it into a technical requirement. Then, just as importantly, that person needs to explain the limitations of the resulting model back to the scientists and leadership team.
That ability is hard to capture in a job description. It is also why I would be cautious about screening candidates based mainly on years of AI experience. Five years focused on building machine-learning systems in an unrelated industry does not automatically make someone better suited to biotech than a computational biologist who has spent three years applying machine learning directly to biological problems.
A few practical questions can help identify the right candidate for the job. What biological data have they worked with? What did they personally build? Did scientists actually use the output to make decisions? How did they validate it? What happened when the model was wrong? Have they built something reproducible, or only a research prototype? Can they communicate with both a research scientist and a software engineer? The answers to those questions will tell you much more than a reference to AI on someone’s résumé.
Companies should not assume every capability has to come from an external hire, as some of the best talent may already be inside the organization. A computational biologist already on staff who develops deeper machine-learning skills may ultimately be more useful than a newly hired generic AI specialist who needs years to develop domain knowledge. The same can be true of a company’s scientists who learn how to design AI-enabled workflows or engineers who develop a serious understanding of the biology behind the systems they are building.
AI is not simply creating another department inside biotech. It is changing what companies look for in scientists, engineers and technical leaders. The organizations that handle that transition well will probably not be the ones that hire the greatest number of “AI people.” They will be the ones that find and develop people who can connect the technology to the science in a way that produces better decisions, better experiments and, ultimately, better medicines.
