We are looking for AI engineers who want to cure diseases — not just write code.

About the role

We are building a control room for disease reasoning—integrated modules that perceive biology, propose interventions, and reason from mechanism to patient impact. This platform accelerates scientific discovery, anticipates disease, and personalizes treatments to profoundly transform human health.

Platform Overview

• Disease State Map: Active canvas predicting downstream effects with mechanism-aware pathway impact and confidence trails. Layers include pathways, proteins, cells, genomics, patient state, tissue inflammation, and candidate states.
• Patient Cohort View: Aggregates data from cohorts (e.g., n = 1,204) to inform therapeutic hypotheses.
• Therapeutic Hypothesis Engine: Templates intervention logic, predicts outcomes, and prioritizes promising therapeutic options.
• Perturbation Reasoning: Explores perturbations to predict mechanism-aware effects and downstream consequences.
• Modality Pathways: Maps therapeutic approaches including antibodies, small molecules, RNA, protein engineering, cell-state modulation, and combinations.
• Evidence Graph: Connects literature, datasets, and experiments with confidence scoring and provenance.
• Human Translation: Bridges mechanisms to patient relevance through biomarkers, endpoints, experiments, and real-world data.

Hypothesis to Proof Workflow

• Understand: Define the biological question and context.
• Hypothesize: Generate mechanism-grounded hypotheses.
• Perturb: Design experiments and in silico screens.
• Design: Prioritize and refine the best interventions.
• Validate: Test, confirm, and analyze outcomes.
• Prove: Advance with confidence to patient impact.

Core Pillars

• Pillar 01: Disease intelligence before drug design. Map disease states, biological drivers, patient subtypes, and intervention opportunities to understand where intervention could matter.
• Pillar 02: From biological complexity to therapeutic hypotheses. Translate disease-state understanding into structured hypotheses: what to perturb, why it matters, expected results, and plausible therapeutic modalities.
• Pillar 03: Designed for validation, not just prediction. Each hypothesis must progress toward experimental and translational proof, including biomarkers, assays, model systems, patient-relevant endpoints, and evidence packages.