I’m working with a well-backed, early-stage company in London building personalised cancer vaccines.They treat this as a deep technical challenge first: modelling the evolutionary arms race between tumours and the immune system, where signals are sparse, ground truth is costly, and predictions need to be trustworthy in the clinic.Small, driven team + real patient projects and data from day one - no waiting around.They’re hiring a Founding ML Engineer to own and build the core ML platform that designs these vaccines.The problem space is unusually interesting:Sequence models + structured prediction for neoantigen targetingEvolutionary dynamics, selection pressures, inference from limited/noisy observationsGame-theoretic thinking about tumour-immune interactionsYou’ll go from research ideas → production systems: standing up training infra, leveraging serious compute, prioritising clean abstractions.They want someone who:Has strong mathematical foundations (reads methods sections in top papers comfortably)Brings a genuine publication record in strong venues (NeurIPS, ICLR, ICML etc.)Can independently build foundational models - mix of fine-tuning SOTA architectures + designing from scratchIs passionate about hard, high-stakes prediction problems (sceptical of pure leaderboard chasing)Can ship fast without regrettable tech debtBiological domain knowledge helpful but not required - they’ll teach the oncology context.Why this role stands outBacked by genuine industry leaders + top US VCsStrong initial funding and long runwayLive customer projects - impact from week oneLarge R&D budget + fast experimental validation loops for modelsFlexible start dates (finishing PhD later in 2026? No issue)Hands-on founding role with real ownership. London-based.Salary: £100,000 - £200,000+ meaningful equity (they pay for impact).Central London (On-site)