StealthCo is a seed-stage techbio company, created within DSV, building a computational drug discovery platform that constructs causal biological networks from large-scale primary human single-cell omics data and structured published experimental literature. Our multi-agent AI system reasons over these networks to generate, simulate, and rank mechanistic hypotheses for combination therapies — with system accuracy verified against top-tier researchers at the Allen Institute. We will initially focus on oncology indications.

The Role (remote, timezone-restricted)

You will design and build production bioinformatics pipelines for new modalities—spatial transcriptomics, single-cell proteomics, and spatial proteomics—extending our existing scRNA-seq infrastructure. These pipelines feed directly into an agentic hypothesis generation system: the quality of what goes in determines the quality of every therapeutic hypothesis that comes out.

You’ll work closely with our Head of AI & Technology (Dr. Francesco Moramarco) and Head of Platform (Dr. Moustafa Khedr) to:

• Build end-to-end pipelines (ingestion, QC, normalisation, integration, annotation, differential analysis)
• Design modality-specific statistics: spot deconvolution, spatial autocorrelation, ADT normalisation, protein-RNA joint embedding, segmentation, spillover correction
• Extend hierarchical cell type annotation across modalities
• Codify best-practice workflows into reusable templates for agent execution
• Sanity-check outputs to catch batch effects and artefacts before they propagate

Requirements

• PhD in computational biology, bioinformatics, genomics, systems biology, or related quantitative field
• 2–6 years experience in early-stage/high-growth startups
• Pipeline-building experience with spatial transcriptomics (Visium, MERFISH, Xenium) from scratch
• Experience with single-cell or spatial proteomics (CITE-seq, CyTOF, CODEX, IMC)
• Strong Python engineering in the anndata ecosystem (scanpy/squidpy/muon)
• Deep single-cell & spatial statistics knowledge (pseudobulk, multiple testing correction, mixed-effects models, compositional analysis)
• Strong biology grounding; can distinguish biology vs confound; assess mechanistic plausibility
• Timezone: at least 5 hours overlap with UK working hours (UTC−4 through UTC+4 preferred)

Strong desirables

• Tumour biology / cancer immunology (TME, immune evasion, resistance)
• Comfort working in an AI-mediated workflow and writing analysis plans executed by agents
• Experience building pipelines/tools consumed by others; cloud compute (GCP preferred); R proficiency

Nice to have

• Wet lab experience and familiarity with the 10x Genomics ecosystem

We know job descriptions like this can read as a wish list. If you don't tick every box, but believe you can build what we need - apply anyway! We care more about what you've built and how you think than whether your CV maps perfectly to every bullet point.

Benefits

Competitive compensation commensurate with experience and profile, plus equity participation. Flexible arrangement depending on location and preference (full-time employment or long-term consultancy). Small, technically intense team with high autonomy and ownership. Remote-first. Minimal management layers and direct impact on decisions.