COMPANY

We started with one failedexperiment and haven't stopped

A decade later, we're still building the AI infrastructure that drug discovery has been missing — so good science never gets lost to a system that should know better.

Our Origin

In 2015, a PhD student named Tom Leung lost rare patient samples because of a bad antibody. It wasn't carelessness — it was a systemic failure. The scientific evidence to choose the right antibody existed somewhere in the literature. Tom just couldn't find it in time. No one could.

The UHN Lab where the BenchSci origin story began. — fig.I, UHN Lab

Tom reached out to machine learning experts David Chen and Elvis Wianda at the University of Toronto with a question: could AI be trained to read scientific papers the way a PhD scientist reads them — and extract the experimental evidence buried inside? They built a prototype. It worked.

Liran Belenzon joining BenchSci as co-founder and CEO. — fig.II, Liran joining as co-founder and CEO

None of them had ever built a company. Liran Belenzon, an MBA candidate working at the Creative Destruction Lab, cold-messaged them to apply to the accelerator. A few weeks in, he joined as co-founder and CEO.

Four people — three PhDs and an MBA — with one conviction:
the biggest bottleneck in drug discovery wasn’t scientific talent. It was scientific infrastructure.

That conviction has driven every decision we’ve made since.

Timeline

2015
2017
2019
2022
2023
2025
Today

The Creative Destruction Lab (CDL)

Founded at the Creative Destruction Lab, University of Toronto. Four co-founders. One prototype. A mission to stop good science from being lost to bad infrastructure.

First Product Launch

First product launched: AI-Assisted Antibody Selection — the first AI that could read a scientific paper and understand the experiment being described. Scientists at leading academic institutions started relying on it to reduce failed experiments. The problem was bigger than antibodies.

Investments

Expanding to the enterprise. We extended the technology across all reagent and model system selection. Sixteen of the top 20 pharma companies started using BenchSci. Over 50,000 scientists across 4,500 research centers worldwide. Google's AI fund, Gradient Ventures, took notice early. F-Prime and Inovia Capital followed.

Project Butterfly

Working in partnership with global pharma R&D leaders, we turned our decade of scientific data into something nobody had built before: the world's first evidence-backed map of disease biology. Every relationship between a target, a pathway, a disease, and a mechanism — extracted, structured, and connected. Inovia Capital and TCV backed this vision with a $50M Series C.

ASCEND and Series D

We launched ASCEND — giving scientists an unprecedented understanding of disease biology across all therapeutic areas. Not a search tool. A reasoning platform. Generation Investment Management led a $70M Series D, bringing total investment to over $200M. The thesis: this isn't a productivity tool. It's infrastructure for how drug discovery gets done.

EMET: The Agentic Workbench

A decade of proprietary data. A decade of working alongside PhD scientists in biopharma. A decade of building the ground truth of disease biology. We brought it all together into EMET — the agentic research environment that reasons and acts like a team of PhD scientists, purpose-built for preclinical R&D. Just as Cursor changed how engineers work, EMET changes how scientists work.

Evolution

The product has evolved. The conviction hasn't: the biggest thing standing between a scientist and a breakthrough is infrastructure. We're here to remove it.

Our Mission

Our mission is to unravel the complexity of disease biology — so that fewer drugs fail, fewer years are wasted, and more patients get the medicines they need faster.

We are a science-first, technology-deep company. PhD scientists and engineers work side by side here, because we believe the only way to build AI that earns a scientist’s trust is to deeply understand what scientists actually do. We serve biopharma R&D only. We go deep, not wide.

And we measure ourselves by one thing: whether the science moves faster because we exist.