This case study highlights the use of the Erasmus platform to run searches for specialist talent within a technology vertical. In this instance we focus on talent with overlapping skills in protein engineering and machine learning models, as these skills have large applications in biotech.
Protein engineering as a subfield has grown consistently over the last 20 years. However, new skills focused on protein design using machine learning or artificial intelligence has now placed this bio x AI as a fast-growing area of new applied research.
Protein engineering as a subfield has grown consistently over the last 20 years. However, new skills focused on protein design used alongside machine learning or artificial intelligence this output jumps up once more as this segment is a fast-growing area of new applied research. This highlights the importance of boundary conditions and key words, when searching as scientific disciplines are constantly evolving and creating new language to describe bleeding edge research.
Experience in Bio x ML Talent Pool
AI was used in 45% of UK protein engineering research papers published in 2025, compared to 22% in 2019. The number of scientists working on this research in the UK between 2015-2019 was seen to be 38K people, which was now risen to 45K people in 2020-2025. Of these, there is a core pool of talent (~18K people) who were active before and after the mass roll out of machine learning-powered tools for protein structure prediction and design.
Using our platform, we can begin to estimate the experience of these key talent pools. Most talent sits within the range of 5-10 years’ experience, signalling that there are likely many skilled individuals taking up positions in industry or as postdoctoral researchers. Despite this, AI skills are now widely distributed across the protein engineering talent pool with the highest proficiency being 2/3 of the 1-5 years’ experience cohort.
Map of UK Talent Hotspots in protein engineering x ML
Affiliations give us an indication of where this talent is distributed within the UK. While these researchers are mostly concentrated to the London, Cambridge and Oxford, there are other local hotspots in Edinburgh, Manchester and Nottingham. However, there is still a gap between academic and industrial adoption with 187 universities publishing protein engineering using machine learning, but only 23 companies, signalling that the technology is still maturing and requires optimisation before large-scale spin-out activity.
Revena’s Conclusions
- The active UK talent pool in protein engineering is estimated to be 45K researchers, with 9K demonstrating experience in applying machine learning.
- Since 2020, AI has become firmly embedded in protein engineering workflows and its use has increased substantially despite a dip in protein engineering papers.
- The most populous talent cohort is the 5-10 years’ experience range. However, the incoming talent from 1-5 years’ experience show the highest adoption of AI in their workflows.
- Institutions in the UK are forming academic hubs for talent to mature, which will follow on to more talent becoming available for industry recruitment.
Get in touch
If you are interested in knowing more – please contact sam@revena.co.uk