France has one of Europe's most ambitious life sciences ecosystems. Deep scientific talent, world-class research institutions, and a growing generation of biotechs turning discovery into impact. The 4th Annual Paris Digital Science & Innovation Day is where that community comes together to share what's working, what isn't, and what comes next.
This year's agenda is built around the questions we hear most across French and European biopharma: how do you build data foundations and automated workflows that make reliable science possible at scale? How do teams move from digitisation to real operational impact? And what does responsible data governance look like as the tools evolve?
Across four sessions covering biopharma innovation, new modalities and precision medicine, automation and integration in R&D, and process development and scaling, you'll hear from scientists, operators, and R&D leaders sharing real challenges and concrete lessons from the field.
Join 250+ biopharma R&D leaders, data and IT experts, and scientists from across France and Europe for a full day of case studies, expert talks, and panel discussions.
The event is free to attend, but registration is required due to limited capacity.
Sajith Wickramasekara, Founder & CEO, Benchling
Few people sit at the intersection of frontier AI development, pharma’s shifting workflows, and what works in practice in biotech R&D. In this keynote, Benchling CEO and co-founder Sajith Wickramasekara cuts through the noise with a candid view from inside the labs of 1,300+ companies, from large pharma to AI-native biotechs. Sajith will share where AI is delivering in R&D today and why scientific agents haven't taken off like AI coding agents have in software. He'll make the case for an AI Scientist built at the seam between the digital and physical worlds — and why biopharma, of all industries, has the most to gain from AI by acting now.
Simon Turner, Partner - Digital Medicine, Sofinnova Partners
Walid Kamoun, Global Head of R&D Oncology, Servier
Last year, Simon Turner opened with a candid verdict on where biopharma's first wave of AI investments had (and hadn't) delivered. This year, he's back with someone who's been in the trenches.
Walid Kamoun is Global Head of R&D Oncology at Servier, where he has spent the past four years leading one of Europe’s most active oncology pipelines with a clear and unwavering focus on delivering transformative clinical benefits to cancer patients. A key driver of this success has been a deep, pragmatic AI and digital transformation—far from the PowerPoint narrative—embedded in how decisions are made, how science is advanced, and ultimately how therapies reach patients faster and more effectively. A scientist with a background spanning medicine, biology, and computer science, Walid trained at Harvard Medical School under Rakesh Jain and has held R&D leadership roles at Takeda and Merrimack before joining Servier at a pivotal moment for the group, bringing a consistent focus on turning innovation into real, measurable patient impact.
This is a no-slides, no-script conversation about what it actually takes to make AI work inside a large pharma: the trade-offs, the failures, and the few things that genuinely moved the needle. Expect them to get into:
• The experiments that worked, and the ones that didn't. Where AI has genuinely accelerated oncology R&D at Servier, and where it fell short of the hype.
• Data foundations vs. data theatre. What "getting your data house in order" actually looks like when you're not on a conference stage.
• Build, partner, or get out of the way. How Servier thinks about working with AI-native companies like Owkin, AITIA, and In Silico Medicine, and what makes a partnership actually productive.
11:10 — 11:30, Mihir Trivedi, Head of Scientific AI, Benchling
Scientific AI for Designing Better Molecules
In this session, Mihir will dive into the history of AI-assisted drug discovery and the advancements that are accelerating change in this space. He'll cover the explosion of recent AI and ML models for drug discovery, and take a look at how Benchling has been integrating these models into the platform, from both a technical and product perspective.
11:35 — 12:05, Panel Discussion
Panel: AI by design - What does it really mean to build a biotech on AI?
Adèle James, Co-founder & CTO, Phagos
Dina Zielinski, Principal Scientist, WhiteLab Genomics
Moderator: Mihir Trivedi, Head of Scientific AI, Benchling
Phagos combines microbiology and AI to design personalised phage treatments that adapt alongside evolving resistant bacteria, and became the first company authorised to market personalised phage-based veterinary drugs in the EU. WhiteLab Genomics uses AI to accelerate and de-risk the development of genomic medicines across gene, cell, and RNA therapy.
In this panel, Phagos and WhiteLab Genomics will get into what building AI-native science organisations actually requires, including:
• What it takes to generate the data that makes AI work
• Where AI has genuinely changed what's scientifically possible — and where it hasn't
• The regulatory and commercial realities of being AI-first in biotech
12:30 — 13:00, Roundtable discussions (sign up onsite)
Roundtable 1: Two Labs, One Problem: A Wet Lab Scientist's View on the Data Divide
Data scientists need clean, structured outputs. But what does that demand mean for the scientists actually generating the data? Join this roundtable to share experiences from the bench — the workarounds, the pressures, and what it would actually take to bridge the gap between wet lab reality and data team expectations.
Roundtable 2: IT & Digital Strategy: Tackling Infrastructure & Data Challenges
Are you in IT or Digital Innovation? Connect with fellow IT leaders to discuss architecture, data integration, security, and the technical hurdles of modernizing life science environments.
Roundtable 3: Benchling Power Users: Tips, Tricks, and Best Practices
Are you an active Benchling user? Gather with other Benchling users to exchange user insights, share configuration best practices, and discuss how you are overcoming daily adoption or workflow challenges.
12:30 — 13:10, Hands-on AI for Scientist Workshop* (limited to 32 seats)
AI is already delivering real results in biotech R&D. The challenge for many teams is not whether it works, it is knowing where to start. Using examples and prompts from the Benchling AI team, this workshop will show how scientists are applying AI across their day-to-day workflows, without compromising scientific quality.
In this session we will cover:
• Turning hours of data entry into minutes
• Generating reports and experiment documentation automatically
• Surfacing insights hidden across years of institutional knowledge
• Data conversion, literature synthesis, internal knowledge search, and report generation
You will walk away with practical, copy-paste prompts you can use the same day, a clear understanding of how teams are enabling confident, scalable AI adoption, and the tools to spend less time on busywork and more time on science.
*The workshop is now at full capacity. If you're on the waitlist, we'll be in touch as soon as a spot becomes available. For any questions, please contact himani.gupta@benchling.com.
13:30 — 13:50, Kris Modig, Scientific Solutions Consultant, Benchling
Benchling Biologics: AI Infrastructure for Modern Antibody Discovery
Most data infrastructure wasn't built for modern antibody discovery; systems can't keep pace with novel formats, making it hard to query, analyze, or train models on antibody data. To solve this problem, we're introducing Benchling Biologics: a purpose-built system for capturing antibodies with full structural awareness. We'll show you how teams can rapidly design custom formats, from mAbs to complex multispecifics, and register proteins with built-in annotation and validation. Because Biologics is native to Benchling, sequence data and experimental context stay connected across the design-build-test-learn cycle.
13:55 — 14:15, Jacques Fieschi, CSO, MImAbs
De-Risking Therapeutic Antibody Development: Biological Tools in the Age of AI
MImAbs delivers end-to-end antibody discovery and ADC development through a unique combination of single B-cell functional screening (Beacon®), high-throughput bioproduction, and expert conjugation chemistry. Their function-first approach enables direct selection of therapeutic candidates based on native cell behaviour — affinity, internalization, neutralization, and cross-reactivity — in a single workflow.
In this session, Jacques will share how predictive computational tools offer growing opportunities to improve candidate pre-filtering, and conjugation rationalization — compressing timelines without compromising biological relevance. As AI reshapes antibody discovery, MImAbs is positioned at the interface: combining irreplaceable biological insight with emerging algorithmic strategies to accelerate therapeutic antibody development.
14:20 — 14:40, Jean Louis Honeine, Solution Consultant - IT & Data Science, Benchling
From Pipette to Pipeline: Automating the Data Foundation for AI in Pharma
Everyone is talking about AI in pharma. Few are talking about what makes it work: a continuous, reliable supply of clean experimental data. This talk makes the case that lab automation isn't just an efficiency play — it's the critical infrastructure for any serious AI strategy. Drawing on Benchling's bulk plate creation and high-throughput cell automation capabilities, we'll walk through how automated workflows reduce human error, enforce data standards at the point of capture, and feed the structured datasets that machine learning actually requires. Practical, no hype.
14:45 — 15:15, Panel Discussion
Baudoin Delépine, Head of Computational Biology, Generare
Julien Duquesne, CTO & Co-founder, Scienta
Moderator: Jean Louis Honeine, Solution Consultant - IT & Data Science, Benchling
The Data Foundation: Structure, Standardisation and Scale
The most important decisions in AI-driven drug discovery aren't made when you train a model — they're made when you design your datasets. In this panel, Baudoin Delépine (Head of Computational Biology, Generare) and Julien Duquesne (Co-founder & CTO, Scienta) break down what that means from two very different vantage points. Baudoin will share how Generare structures the computational biology pipeline behind its microbial molecule discovery platform — triaging genetic recipes at scale and turning raw sequencing data into something a drug developer can act on. Julien will speak to the data challenges behind EVA, Scienta's foundation model for immunology and inflammation — specifically what it takes to harmonise transcriptomic data across species, platforms, and resolutions before a single prediction can be trusted. Two companies. Two data problems. One shared conviction: the foundation has to be right before the science can scale.
15:45 — 16:05, Gustave Ronteix, Co-founder and CTO, Orakl Oncology
The challenges of high-throughput data generation for training AI perturbation models
Perturbation models follow scaling laws. They require large functional datasets pairing drug-response phenotypes with multi-omic backgrounds, and these barely exist today. Gustave Ronteix, Co-founder and CTO of Orakl Oncology, will share how the team is tackling this bottleneck directly. The presentation will cover:
• How Orakl cultures patient-derived lines at scale, running systematic drug panels across chemotherapies and targeted therapies to build a high-quality training corpus
• Enabling the lab-in-the-loop flywheel: building the engineering systems and processes to upload lab data at scale, and automatically correct and QC large volumes of sophisticated biological data on the fly
• A case study in data quality: spotting pipetting errors in real time and eliminating the 1% mistakes that corrupt training sets
16:10 — 16:30
Building digital foundations for scalable process development
Speaker & abstract tba
16:35 — 16:55, Jason Boyd, Head of Regulated Labs Market Strategy, Benchling
From Fragmented to AI-Ready: Building the Digital Foundation for Modern Process Development
The biopharmaceutical industry is shifting to unified, data-driven process development with digitization and AI, replacing manual workflows. Organizations must design intentional digital workflows for AI-ready data, boosting efficiency and collaboration. Moving beyond fragmented systems to a cohesive digital foundation is key. Benchling drives this modernization with a purpose-built, process-centric data model, ensuring structured, connected, and actionable data. This allows teams to move faster, collaborate better, and use AI to accelerate drug development.
Jinel Shah, Scientist, Cradle Bio
Laura Lloyd, Life Sciences R&D Global Account Lead, Cognizant France
Nathalie Anthoine, Digital Health & Science Lead, Opella
Moderator: Meritxell Orpinell, Head of CX EMEA, Benchling
Women in Biotech is an inclusive panel series hosted by Benchling that invites open, honest dialogue about leadership, identity, and growth for women in the life sciences. This session creates space for diverse voices to reflect on their journeys in science and technology—where they've come from, what they’ve overcome, and how individuals envision their future in a rapidly evolving landscape shaped by AI. Together, we’ll challenge outdated norms, surface unspoken challenges, and uncover new paths to sustainable, authentic leadership.
Learn more about Benchling at benchling.com
