BenchSci

Six sharp takeaways from our SVP of Strategic Alliances, Casandra Mangroo, on what the Pistoia Alliance Spring Conference said out loud about where pharma AI is heading. The closing question is the one to sit with: "Are you building a faster car, or reimagining the engine?" 👇

Drug discovery moves faster when the industry moves together. We’re proud to join the Pistoia Alliance, a global network of pharma, biotech, and technology leaders working together to advance life sciences R&D. Through this partnership, we’re excited to collaborate more deeply with industry peers, contribute to advancing data and AI/ML practices, and continue learning from the broader ecosystem. Because real progress in drug discovery happens when we build—and learn—together.

What does it take to build a company tackling one of the hardest problems in R&D? In this episode of Born in Silicon Valley from Match Relevant with Jake Villarreal, our CEO, Liran Belenzon, shares: • Why most drug failures come down to misunderstood biology • How BenchSci acts as an AI co-pilot for scientists • What 10+ years of building teaches you about focus, scale, and resilience This is a candid look at science, leadership, and decision-making in drug discovery. 🎧 Listen here: https://lnkd.in/eFbAmwJq

“Inference will become a reagent.” This shift will redefine how drug discovery actually gets done. Our CEO, Liran Belenzon, shares a perspective on what this looks like in practice—and why it matters now.

Drug discovery is entering a new era—and the shift is already underway. In his latest piece, Liran Belenzon breaks down five forces reshaping the future of how we build new therapies. For anyone building, investing in, or working at the intersection of AI and life sciences, this is a must-read.

Undisciplined AI tools often increase friction rather than speed in preclinical R&D. While many organizations attempted to replace human expertise with unguided AI agents in 2025, the results were frequently erratic. In the context of drug discovery, "black box" engineering is a significant liability. Accuracy is not optional when failure has real-world consequences for patient outcomes and IND timelines. At BenchSci, we address this through a practice we call AIDE (AI-Assisted Development and AI-Driven Engineering). This disciplined approach ensures our neurosymbolic architecture powers ASCEND with the scientific rigor you require to: - Eliminate redundant experiments through evidence-based reasoning. - Maintain data integrity across complex, multimodal datasets. - Transition from experimental AI to a reliable partner in scientific discovery. In our latest technical deep dive, Principal Engineer, Amit Bronner, shares how we are architecting AI preparedness for the enterprise 👉 https://lnkd.in/g3fSF9n3 #DrugDiscovery #PreclinicalResearch #AIEngineering #Biotech #BenchSci

Yesterday’s Bench to Industry event brought together scientists, industry leaders, and curious minds exploring what impact beyond the bench can look like. 🧬 The energy? High. ⚡ The questions? Thoughtful. 🤔 The career paths? Anything but linear. 💡 From honest reflections on career transitions to practical advice for navigating industry, the room was filled with thoughtful questions and generous insights. A huge thank you to our incredible panelists, Luca Delfinis, PhD, Ran Antes, Ph.D, Elissa Currie, and Luís Eduardo Abatti, PhD, who shared their experiences so candidly and helped make the conversation so meaningful. 👏 And the biggest takeaway: There’s a growing community of scientists reimagining where their skills can go next. 🚀

Last week, our team was in Boston for Lab of the Future, where we had the opportunity to connect with scientists, innovators, and partners from across the life sciences ecosystem. One of the highlights was Casandra Mangroo’s keynote, where she shared how our collaboration with Thermo Fisher Scientific will bridge the gap between the dry lab and the wet lab. Thanks to everyone who stopped by to connect with our team during the conference.

Biomedical researchers are drowning in scientific data. The challenge isn’t just volume — it’s that biological knowledge is fragmented across millions of papers, datasets, and experimental contexts. A new case study from Neo4j explores how BenchSci is tackling this challenge. The piece highlights how our ASCEND platform uses a neurosymbolic AI approach — combining LLMs with a structured Biological Evidence Knowledge Graph (BEKG) — to decode, harmonize, and connect biomedical evidence. The result is a computable map of disease biology, built from tens of millions of publications and datasets. By modeling the relationships between genes, diseases, experiments, and outcomes, researchers can generate stronger hypotheses while ensuring every insight is traceable back to experimental evidence. The case study features BenchSci's VP of Engineering, Machine Learning, & Data, Tony Solon PhD, who shares why graph technology is critical for representing the complexity of biology at scale. Read the full case study 👉 https://lnkd.in/gtHYdwxF

🚀 Only a few spots left for next week's event, 'From Bench to Industry'! RSVP now to secure your spot 👉 https://lnkd.in/evvRh7_A Thinking about a move from academia to industry? Curious what that transition actually looks like beyond the job titles and LinkedIn headlines? Join us for BenchSci Presents: From Bench to Industry, an in-person panel discussion featuring scientists who’ve made the leap—and are ready to share their experience. 💬 Expect honest conversations about navigating the job search, translating academic skills into industry impact, adapting to new cultures, redefining success beyond the lab, and plenty of time for audience Q&A. 🎤 Meet the panel: ✔️ Elissa Currie, Ph.D, Senior Product Marketing Manager, BenchSci — PhD-trained scientist whose interest in the intersection of science, AI, and technology evolved into a role in product marketing driven by storytelling and data-informed strategy. ✔️ Luís Eduardo Abatti, PhD, Science Tech Lead, BenchSci — From wet-lab cancer research to building AI systems that decode scientific data. ✔️Ran Antes, Ph.D, Science Manager, BenchSci — Leads scientific quality and evaluation for BenchSci’s conversational AI platform, turning scientific data into reliable, production-ready products. The discussion will be moderated by Luca Delfinis, PhD, Scientific Liaison at BenchSci—a scientist specializing in preclinical therapy development who now works closely with pharma R&D scientists to better understand and support their research needs. 📅 Tuesday, March 10, 2026 ⏰ 6:00–8:00 PM ET 📍 BenchSci Office | 559 College St, Suite 201 ⚠️ Space is limited—RSVP here: https://lnkd.in/evvRh7_A