Proteins power life, yet most of their potential remains untapped. At CureCraft, we fine-tune large-scale protein language models and couple them to bespoke evaluation engines so we can design, rank, and iterate on novel enzymes entirely in-house —from GPU to benchtop, under one roof. We are a venture-backed and high-profile team. With our own wet-lab facility and a growing proprietary dataset, we’re ready to accelerate the loop between in-silico ideas and in-vitro proof. Your work will directly translate into engineered proteins with improved stability, activity, and entirely new functions. What you’ll do Invent & implement next-gen architectures for protein sequence, structure and fitness prediction (transformers, diffusion, hybrid approaches). Close the loop between model output and experimental data—design active-learning or RL-based strategies that get smarter with every assay. Scale the pipeline: own data engineering, model training & evaluation on multi-GPU/TPU stacks; productionise successful models for routine design tasks. Collaborate deeply with bioinformaticians and wet-lab scientists; turn noisy experimental readouts into actionable signals. Stay ahead of the curve: track and selectively apply cutting-edge advances from GenAI, computational biology and adjacent fields. Contribute to culture: mentor junior engineers, shape best practices, and help us build a genuinely interdisciplinary, mission-driven team. About you PhD (or equivalent experience) in Machine Learning, Computational Biology, Bioengineering, or related field. Hands-on mastery of deep-learning frameworks and modern generative modelling. Proven track record applying AI to protein or molecular problems —publications, open-source contributions, or industrial success stories. Strong grasp of statistics, experiment design, and the realities of biological data (noise, bias, scale). Ability to thrive in a fast-moving startup: you turn ambiguity into roadmaps and ship. Plus, if familiar with: UniRef, UniProt, MGnify — for pretraining corpora | PDB, AlphaFold DB, FoldSeek — for structure-grounded learning | MSA generation tools — HMMER, HHblits, MMseqs2 | Next-generation Sequencing (NGS) data What we offer High ownership & impact: your models go from idea to bench within weeks, not years. Top-tier compute, lab resources and an annual conference/learning budget. Flexible hybrid schedule from our lab/office in London. A tight-knit, curious team that cares about doing science that matters. How to apply Send a short note ( jobs@curecraft.co with the subject line “Senior AI Scientist – [Your Name]”. Links to publications, code, or relevant projects are warmly welcomed. Applications are reviewed on a rolling basis—reach out soon if you’re ready to craft the next generation of cures.