ArcaScience

One drug, 67 PICOs. The evidence map your BRA can't afford to miss A new peer-reviewed analysis just quantified something the JCA community has been sensing for years: PICO scoping under the EU HTA Regulation is far more complex than most HTDs are prepared for. Key findings from 35 multi-country exercises: -Average 8 consolidated PICOs per indication -A single NSCLC asset generates up to 67 PICOs, driven by biomarker cut-offs and histology sub-populations -Most published exercises covered only 3–10 countries; JCA involves all 27 -HTDs have 100 days to finalize the dossier once scoping closes -Consolidation criteria remain opaque, no clear published thresholds for which PICOs get merged or dropped Evidence generation planning needs to start much earlier than most teams realize. Full analysis: https://lnkd.in/eAHr9fFK #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #HTA #JCA

Tune into our CEO Romain CLEMENT, feature in a vidéo in partnership with Future4care. In healthcare and pharmaceutical development, time is not just a matter of cost, it is a matter of lives and access to treatment. ArcaScience is redefining how leading pharmaceutical companies assess the benefit-risk profile of their drug candidates. Using our specialized AI models, we enable the industry to uncover the true mechanism of action of a drug long before it reaches patients. Our technology is designed to transform the R&D pipeline: - Unmatched precision: Our clients uncover 90% to 99% of critical insights that would otherwise remain hidden, leveraging 9 to 100 times more relevant data. - From clinical trials to post-market surveillance: Our solution supports decision-making throughout the entire drug lifecycle, from clinical development to post-market monitoring, allowing the benefit-risk profile of treatments to be continuously refined. - Full transparency: No more AI "black boxes." We provide clear, explainable access to every insight and the underlying data. Our mission is to help develop safer medicines, faster, and ensure they are better tailored to patients' real needs. #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup #healthtech #future4care

Across the industry, teams are constantly looking to make high-stakes decisions such as which drug to advance, what trial to redesign and which signal to trust, all based on fragmented, partially integrated evidence. By the end of 2026, 80% of pharma companies are expected to rely on AI for benefit–risk evaluation. This is a literal shift in infrastructure. At the same time drug development costs continue to rise, late-stage failures remain the norm and timelines in the drug development industry are increasingly under pressure. As a result, we are still paying for uncertaity instead of investing in understanding. At ArcaScience, we see AI not as a layer on top of this system, but as the infrastructure to change it. AI applied to drug discovery is becoming a way to : - Integrate evidence prospectively, - Structure benefit–risk earlier, and - Make decisions with clarity, not hindsight The question is no longer if this shift will happen. It’s already happening. The real question is: Who will adopt it early enough to turn uncertainty into advantage? #AIAccountability #BenefitRiskAssessment  #DrugDevelopment #HealthcareAI #Pharmacovigilance #frenchtech #startup #healthtech #biotech #techbio

If an AI system contributes to a medical decision, who is responsible? The clinician? The institution? The developer? The regulator? As AI becomes more embedded in drug development and clinical strategy, this question is no longer theoretical. AI does more than generate insights: it influences decisions, sometimes at scale. In healthcare, decisions have consequences: • For patients • For clinical outcomes, and • For entire development programs. Rather than rejecting AI, we have to avoid systems where: • Reasoning is opaque, • Assumptions are hidden, and • Accountability is unclear. At ArcaScience, accountability is a core design principle. AI in healthcare must be: • Traceable, • Auditable, and • Explicit in its assumptions. As a requirement, not a feature. Trust in AI won't come from performance alone. It will only come when there is clarity on how decisions are made and who stands behind them. #AIAccountability #BenefitRiskAssessment #DrugDevelopment #HealthcareAI #Pharmacovigilance #ExplainableAI

A short while ago, Vassili Soumelis, CMO of Arcascience was a guest on Med In Tech. In this clip, Vassili dives into a fascinating topic at the very core of tomorrow's medical breakthroughs: the creation and utility of the virtual patient. By combining multiple data sources, whether clinical data, patient-specific characteristics, or medical imaging, it is now possible to achieve an unprecedented level of understanding and representation. The ultimate goal is running real-time simulations on this "virtual patient" to refine diagnoses, anticipate treatment responses, and pave the way for ultra-personalized precision medicine. #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup #healthtech #medintech

In drug development, we often talk about “evidence.” But what does good evidence actually mean? Today, decision-making relies on a patchwork of sources : • Clinical trial data • Scientific publications • Real-world evidence • Safety reports • Mechanistic biology While each of these is valuable, none of them, on their own, is sufficient. The real challenge is the ability to integrate heterogeneous signals into a coherent, decision-ready understanding. Because without integration, strong signals remain isolated and weak signals can be overinterpreted This is exactly the problem ArcaScience aims to address. We believe that “good evidence” is not defined by its source, but by its ability to be : • Structured • Contextualized • Continuously evaluated within a broader model In other words : Evidence only becomes valuable when it becomes actionable. #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup

Glad to be joining France Deeptech  alongside 14 other startups building the future of French tech.  🚀 At ArcaScience, we put AI to work on benefit-risk assessment to make health decisions faster and more reliable. Looking forward to connecting with this ecosystem. Thanks to the France Deeptech team for the welcome!

“Investors don’t want to fund dreams, they want to fund inevitabilities,” explains Romain CLEMENT this week in Link to Leaders. Through this op-ed, Romain shares a conviction shaped by his entrepreneurial journey: true fundraising readiness does not begin with a pitch deck, but much earlier, with the ability to solve a real problem, backed by concrete evidence and measurable impact. Romain also discusses: - the importance of starting from a problem that is deeply understood and personally experienced - the need to build a solution validated by the field before seeking funding - the essential role of discipline, efficiency, and long-term vision in building a sustainable company It's a thoughtful reflection on entrepreneurship and the future of emerging technologies in both Portuguese and English. Thank you to Link to Leaders https://lnkd.in/ets8q8Zw #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup 

This week, StartupReporter sat down with Romain CLEMENT for a five-question interview exploring his entrepreneurial journey, the vision behind the company, and the technological challenges shaping the future of techBio. What inspired the creation of ArcaScience? Why are standard protocols no longer sufficient in precision medicine? How can explainable AI fundamentally transform pharmaceutical R&D? And what will it take to make drug development more predictable, transparent, and personalized? In this interview, Romain shares the personal story that sparked ArcaScience, the lessons learned from building AI for healthcare, and his perspective on the next generation of techBio infrastructure. A big thank you to Startup Reporter for the opportunity to share our vision and discuss the future of AI-driven healthcare innovation ! https://lnkd.in/eFCJEa7W #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup

𝐖𝐡𝐚𝐭'𝐬 𝐚𝐜𝐭𝐮𝐚𝐥𝐥𝐲 "𝐡𝐨𝐭" 𝐢𝐧 𝐭𝐡𝐞 𝐧𝐞𝐰 𝐞𝐫𝐚 𝐨𝐟 𝐩𝐨𝐬𝐭-𝐡𝐲𝐩𝐞 𝐓𝐞𝐜𝐡𝐁𝐢𝐨? After two years of profound market transformation, investors are now backing infrastructure capable of delivering explainable science, high-quality data, and reproducible outcomes. In his latest opinion piece published by EU-Startups, Romain CLEMENT explores the key trends reshaping the TechBio landscape: - The decline of AI "wrappers" and black-box solutions. - The rise of high-fidelity data infrastructure. - The shift from discovery-driven research to predictive engineering. - The growing importance of safety, compliance, and modelling biological resilience. An insightful perspective on where investment is flowing, and why the platforms that can transform complex healthcare data into scalable, actionable assets will define the future of TechBio. Thanks to EU-Startups for the publication https://lnkd.in/eTx3bcJe #techbio #AI #frenchtech #biotech #datascience #clinicaltrials #startup