BioTwin, founded by Louis-Philippe Noel, Pierrick Haugel and Nicolas Antic, is rethinking how healthcare data is used, moving beyond static diagnostics toward continuous, predictive insights. By building virtual patient twins from longitudinal data, the company aims to detect disease earlier, improve clinical decisions, and shift healthcare from reactive treatment to proactive, preventative care.

1. Every startup begins with a moment of insight. What problem or experience first inspired you to build this company?

Medicine excels at producing data, labs, imaging and wearables but it still rarely reads that data as a continuous story. The founding insight behind BioTwin was direct and urgent: biological risk trajectories evolve silently long before symptoms appear, yet clinical practice and most diagnostic tools operate on isolated snapshots. Watching a close colleague receive a preventable late-stage diagnosis crystallized the need to change that paradigm. We built BioTwin to convert years of multimodal biological signal into a synchronized, auditable virtual twin for each patient a living virtual representation that flags deviations early, simulates interventions, and documents every inference for clinical accountability.

2. What is the core problem your company is trying to solve, and why do you believe existing solutions haven’t fully addressed it?

The core problem we solve is structural. Late-stage disease costs materially more to treat, and screening programs miss a meaningful fraction of high-risk individuals. Existing solutions fail because they treat assessment as discrete events: electronic health records accumulate data but do not model trajectories; diagnostics give point-in-time readouts without longitudinal context; precision-omics initiatives produce rich signals that rarely translate into operational workflows clinicians can trust. Three gaps persist at scale: fragmented data across systems, no discipline for temporal synchronization and quality control, and a lack of auditability required for Health Technology Assessment and reimbursement. BioTwin was built to close those gaps in parallel.

3. Your platform sits at the intersection of technology and innovation. Can you explain how your solution works in simple terms and what makes it technically unique?

In plain terms, BioTwin ingests longitudinal signals at-home dried blood spot (DBS) sampling, untargeted metabolomics, clinical records, wearables, and patient-reported data and fuses them into a time-aware virtual twin. That twin is not a static profile: it's a continuously updated patient model that detects early deviation from a patient's baseline, runs bounded "what-if" simulations for clinical decisions, and produces a complete provenance trail for every prediction. Technically, our differentiation is threefold: pre-analytical rigor for DBS collection and sample handling; an RDF-based model registry and canonical fusion architecture that aggregates attribute-level models; and a provenance-first design that makes every inference traceable, versioned and explainable — the bridge from research accuracy to clinical accountability.

4. Can you share a real-world example or use case that best demonstrates the value your technology brings to users or businesses?

We demonstrate value through anchored clinical pilots. Our research partnership with Cleveland Clinic Abu Dhabi (CCAD) focused on oncology screening. In Phase 1, BioTwin produced promising preliminary results across multiple cancer types (preliminary cohort; results undergoing peer review). That early screening potential translates into earlier therapeutic windows, improved curative opportunity, and better treatment economics where a stage shift materially reduces lifetime cost of care. Parallel work includes a Phase 1 Parkinson's pilot where longitudinal metabolomics combined with sensor data refines prognostic trajectories and improves trial recruitment signals — a direct value lever for developers and clinicians running early therapeutic studies.

5. What has been the most difficult challenge your team has faced since launching, and how did you overcome it?

Operationalizing those outcomes required solving a hard technical-operational problem: how to make untargeted metabolomics clinically reproducible. Metabolomics is highly sensitive to pre-analytical variability — collection technique, transit, storage and extraction all matter. We addressed this by containerizing the full analytical pipeline to guarantee reproducible processing across environments and embedding QC gates that reject anomalous data before modeling. Crucially, MBRIF's support was materially important during this phase: their network gave us accelerated access to clinical and industry mentors who helped validate our operational model and stress-test assumptions ahead of prospective trial design.

6. Your company operates in a rapidly evolving industry. What major trends do you believe will shape the sector over the next five years?

Five trends will define the next five years: the shift from snapshot to longitudinal trajectory modeling and corresponding regulatory recognition; the rise of synchronized, federated virtual twin registries enabling population-scale what-if simulations; stricter HTA-grade evidence and health-economic requirements for payer adoption; provenance and governance becoming compliance-by-design; and bounded clinical AI copilots that embed LLM interfaces over validated, auditable models rather than relying on open-ended generation.

7. Looking ahead, what is the long-term vision for the company, and how do you hope it will reshape your industry?

Our long-term ambition is pragmatic and structural: to be the individual-level virtual twin layer that links population health platforms, clinical workflows and payer infrastructure. At scale, synchronized patient twins change incentives from reactive treatment toward proactive prevention by giving clinicians a multi-year, simulation-enabled view before committing to a care path. We are building that from Abu Dhabi and Québec with a clear commercialization path in 2026: expanded CCAD Phase 2 pilots, TwinMe V1 user portal rollout and payer integrations designed around HTA-grade evidence. MBRIF's support was instrumental in accelerating the validation work that now underpins our clinical roadmap — because physiology must be read over time, not in snapshots.