Katalyze AI Announces $10.5M Seed Round

SAN FRANCISCO, CA - Katalyze AI, the agentic operating system for pharmaceutical companies, today announced $10.5 million in seed funding led by Bonfire Ventures, along with Inovia Capital, Ripple Ventures, Alumni Ventures, and leading angel investors including Gokul Rajaram and Farzad Soleimani. Katalyze’s platform lets any scientist, engineer, or analyst in a biopharma company build teams of agents that take on real engineering, scientific, and manufacturing work.

Why Now: The Opportunity to Put Agents to Work in Pharma

The path from discovering a molecule to patient has become unprecedentedly complex, high-stakes, and capital-intensive. As pharmaceutical companies face the patent cliff and multi-decade highs in drug shortages, the pressure to compress timelines from discovery to patient is at an all-time high. The opportunity to use AI is enormous, and Katalyze is built for the hardest engineering and scientific problems in pharma, where being approximately right is worthless and an answer has to be correct every time.

Katalyze: One Agentic Operating System for Pharma

Katalyze is a single agentic operating system for pharmaceutical companies, built to deliver accurate results fast across the regulated span from process development to release. The data behind that work is scattered across plant and lab systems, so Katalyze connects it through MCP, a command-line interface, and prebuilt integrations with the systems teams already run, including MES, LIMS, ELN, historians, and SAP. The result is one source of truth that people and agents can both reason over, where every answer is anchored to the data it came from.

That grounding is what makes the rest possible: results carry the verification and traceability regulated work demands, and agents operate within GxP, data privacy, and data sovereignty requirements rather than around them. The platform deploys in weeks, compressing the timeline from idea to dose and lowering the total cost of commercializing a molecule.

“What really separated Katalyze was that it was built for an enterprise like Sanofi from day one. The ontology layer was already in place. Data ingestion into the intelligence layer was solved. They had the security, the governance, the deployment story, everything we needed to scale this across R&D, not just run a pilot in one corner of the organization,” said Sabya Dasgupta, Global Head (VP) R&D Data Platforms & Products at Sanofi.

At the center is a dynamic context layer for every molecule: an operations-specific ontology and knowledge graph that grounds each agent in the facility’s real data, so every output is accurate and traceable to an immutable source. Those results are sharpened by a community of more than 100 tenured scientists and engineers from companies including Pfizer, Sanofi, and Lilly, who collaborate with Katalyze to author the skills its agents run on, a flywheel that is difficult to replicate.

In one early deployment, an analysis that would have taken a year and $4 to $6 million was completed in 45 minutes.

Key features include:

• Operational Data Layer: Unifies fragmented data across the plant and the lab into a single, real-time operational truth, entirely eliminating manual data assembly.

•  Primary Production Record: Serves as the GxP-native ontology layer, leveraging an operations-specific ontology and knowledge graph to automatically anchor every agent decision and insight back to its immutable source.

•  Katalyze Agent Catalog: Deploys out-of-the-box, domain-trained agents on day one that speak the specialized dialect of MSAT, Quality, and bioprocess engineering to investigate deviations, track CAPAs, and draft APQRs.

• Agent Studio: Empowers internal scientists and data engineers to compose custom, proprietary agents on the same secure, source-grounded substrate, inheriting full regulatory compliance by default.

“The pressure to get medicine to patients faster, and at lower cost, has never been higher, but the bar for accuracy in our industry is absolute,” said Reza Farahani, Co-Founder and CEO of Katalyze AI. “We built an agentic operating system where every answer is grounded in an immutable record, so teams can deploy agents that are right every time and cut lab and manufacturing cycles from quarters to weeks, without ever exposing sensitive data.”

“Most AI in this category is a thin copilot bolted onto legacy tools,” said Brett Queener, General Partner at Bonfire Ventures. “Katalyze went the other way and built real infrastructure. By putting a GxP-native context layer underneath autonomous agents, they let AI reason across the messy, fragmented data of pharma manufacturing and actually do the work: solving deviations faster, improving yield, running critical workflows at a fraction of the cost. That’s the unlock the industry has been waiting for, and Reza’s team is the one delivering it.”

Built by a World-Class Team

Katalyze was built by Reza Farahani (CEO), Shreyas Becker (COO), Hannes Bretschneider (Chief AI Officer), and Matt Cruz (Founding Engineer). The four of them bring decades of experience building startups across AI, enterprise tech, and life sciences. They’ve assembled a deep bench of employees and advisors spanning Sanofi, OpenAI, J&J, and other leading life sciences and AI companies.

With the seed funding, Katalyze will expand its team across engineering, science, and go-to-market, grow its catalog of domain-trained agents, and scale deployments with the world’s largest pharmaceutical companies.

About Katalyze AI

Katalyze AI is the agentic operating system for pharmaceutical companies. The platform unifies fragmented production data into a single source-grounded record to run a catalog of domain-trained agents. Its platform is used by 5 of the 20 largest global pharmaceutical companies. Katalyze is based in San Francisco and is backed by Bonfire Ventures, along with Inovia Capital, Ripple Ventures, Alumni Ventures, and leading angel investors. Learn more at https://katalyzeai.com/.