Biomni Lab Launches with $13.5M Seed
Phylo emerged from stealth on February 3, 2026, unveiling Biomni Lab, what it calls the first Integrated Biology Environment (IBE), alongside a $13.5 million seed round that PRNewswire reported was co-led by Andreessen Horowitz and Menlo Ventures through the Anthology Fund linked to Anthropic. Zetta, Conviction, SV Angel, Valkyrie, and a group of angel investors also participated. The financing gives the South San Francisco startup runway to turn a research project that the announcement stated had already reached 7,000-plus labs into a commercial platform with enterprise-grade security, reproducibility, and dedicated infrastructure.
The product is a workspace where biologists orchestrate AI agents across literature search, experiment design, data analysis, and interpretation, tasks that today live in fragmented scripts, notebooks, and point tools. Biomni Lab integrates over 300 databases, software systems, and analytical tools, plus connections to proprietary providers including Consensus for literature, COSMIC for cancer genomics, and Addgene for plasmid discovery. The new agent architecture, the press release found, outperforms existing agent systems by more than 20 percent on standard benchmarks and long-horizon scientific evaluations. A case study with Ginkgo Bioworks showed the platform accelerating more than ten complex cell-painting and transcriptomic analyses from weeks to hours, with publication-quality results validated by Ginkgo scientists. "Phylo's Biomni Lab platform is user-friendly for all scientists to automate bioinformatics analyses, generate publication-quality figures, and even compare results with external datasets," said Ayla Ergun, Senior Director of Data Science at Ginkgo Bioworks.
The founding team carries unusual depth for a seed-stage company. CEO Kexin Huang, Ph.D., and President Yuanhao (Jerry) Qu, Ph.D., both Stanford doctorates, started the open-source Biomni project in June 2024. Scientific co-founders include Jure Leskovec, Ph.D., Stanford professor of computer science, and Le Cong, Ph.D., Stanford professor of pathology. The advisory bench reads like a who's who of modern biology: Nobel laureate Carolyn Bertozzi, Ph.D.; CRISPR pioneer Feng Zhang, Ph.D.; computational biology pioneer Fabian Theis, Ph.D.; and founding advisor Malay Gandhi, former chief strategy officer at Benchling. The team's prior work spans Biomni, POPPER, CRISPR-GPT, TDC, DeepPurpose, ClinicalBERT, GEARS, and TxGNN, with industry stints at Google, Scale, Factory, Microsoft, Anthropic, Genentech, Pfizer, Benchling, Alector, and Cytoreason.
Investors framed the bet around adoption velocity. "User adoption is one of the strongest signals that we look for," said Jorge Conde, general partner at Andreessen Horowitz. "It's rare, especially in the life sciences, to see an academic research project reach this level of real-world usage and sustained user love." Matt Kraning, partner at Menlo Ventures, added: "Phylo is fundamentally changing the way biological research happens. Kexin and Yuanhao have built breakthrough agentic AI technology and integrated it into a platform that brings AI-native productivity to the lab."
Hiring Sprint Anchors in South San Francisco
The seed round triggered an immediate hiring sprint. The compensation band is $145,000 to $300,000 (median $300,000) for seven salaried roles, including agent engineers, infrastructure engineers, product engineers, forward-deployed scientists, and an enterprise account executive.
| Company | Role | Location | Salary Range |
|---|---|---|---|
| Phylo | Member of Technical Staff – Agent Engineer | South San Francisco | $200,000–$300,000 |
| Phylo | Member of Technical Staff – System Engineering | South San Francisco | $200,000–$300,000 |
| Phylo | Member of Technical Staff – Product Engineering | South San Francisco | $200,000–$300,000 |
| Phylo | Member of Technical Staff – Forward Deployed Scientist | South San Francisco | $170,000–$275,000 |
| Phylo | Enterprise Account Executive | Remote (Toronto) / On-Site (SF Bay Area) | $250,000–$300,000 |
The concentration in South San Francisco is deliberate. Phylo's headquarters sits in the same corridor where Insitro — another AI-native drug discovery player — has posted a Chief Medical Officer role at $450,000–$480,000, a VP of Bioassays at $290,000–$326,000, and a Senior Manager of Imaging Machine Learning at $247,000–$262,000. Phylo's co-founders, Huang, an AI researcher, and Qu, a cancer biology PhD, personify the hybrid profile the company now needs to replicate at scale.
Biomni Lab is not a copilot; it is an Integrated Biology Environment where biologists orchestrate multiple agents across single-cell analysis, human genetics, disease biology, and target evaluation — the exact four workstreams Chugai Pharmaceutical named when it deployed Biomni Lab across its discovery organization in August 2026. Ono Pharmaceutical followed with its own deployment to discovery scientists. The Forward Deployed Scientist role exists to shorten the loop between pharma engagements and platform improvements, placing Phylo talent inside the customer's workflow.
At a $300,000 median, Phylo is pricing its technical roles above the typical biotech software engineering tier and into the range reserved for AI lab alumni. The Enterprise Account Executive band signals that selling agentic AI to pharma R&D now commands a premium over traditional SaaS sales, because the buyer is a computational biology lead who evaluates technical depth.
Phylo's careers page describes the team as "researchers and engineers across AI and biology, commercializing the Biomni platform." The current openings make that description operational. Each role maps to a layer of the agentic stack that did not exist in biotech two years ago. The seed capital paid for the compute and the enterprise hardening; the pharma partnerships proved the workflow; the South San Francisco hiring push is how Phylo turns both into a defensible moat before the next funding cycle.
Two Japanese Giants Bet on Biomni
Ono Pharmaceutical announced its collaboration with Phylo on July 28, 2026. One week later, on August 4, Chugai Pharmaceutical followed with its own deployment announcement. Two Japanese pharma giants, each with decades of drug discovery track records, committed to the same agentic AI platform within days of each other; that is the signal the market needed.
Ono, founded in 1717 and headquartered in Osaka, brings over 300 years of institutional knowledge. Its portfolio includes OPDIVO (nivolumab), the cancer immunotherapy that has reached millions of patients globally. The company's modern strategy explicitly centers on open innovation and AI that shortens the path from idea to novel compound. Seishi Katsumata, Corporate Officer and Executive Vice President of Discovery & Research, put it plainly: "We believe AI will become a core capability for drug discovery. With patients waiting for new medicines, there is an urgent need to help scientists move faster without compromising scientific rigor. Biomni Lab stood out because our researchers quickly adopted it and saw its potential to accelerate everyday discovery."
Phylo CEO Kexin Huang framed the partnership from the other side: "Ono's discovery leadership across oncology, immunology & inflammation, and neurology makes them an ideal partner to demonstrate how this new way of working can accelerate breakthrough research." The language — "everyday discovery," "new way of working" — matters. This is not a pilot tucked into a corner of R&D. Ono intends to embed agentic AI with every discovery scientist.
Chugai's deployment targets the same four workflow areas. That scope suggests Chugai is treating Biomni Lab as a horizontal layer across early discovery, not a point solution for one modality. The August 4 announcement came from Phylo directly, confirming the platform would be deployed "across Chugai's drug discovery workflows."
What Biomni Lab actually does for these organizations: it functions as an Integrated Biology Environment, that workspace now executing end-to-end workflows. The platform synthesizes experimental history, reasons over internal data, designs experiments, and runs computational biology analyses. Phylo claims over 100 pre-built agentic workflows across the biopharma value chain, with deterministic pipelines for reproducibility. The system is model-agnostic, matching tasks to the best available model dynamically. Enterprise-grade security is baked in: ISO27001 certification, SOC2 Type2 compliance, single-tenant cloud environments, and a commitment that customer data never trains the underlying models.
The validation signal compounds when you consider the adoption baseline. Thousands of labs already using Biomni before these enterprise deals create a network effect. Ono's historical data and discovery expertise, combined with Biomni Lab's agentic infrastructure, is explicitly positioned to help scientists move from questions to discoveries in a fraction of the time.
Two partnerships do not prove a category winner. But they establish a beachhead. When a 300-year-old company and a global pharma leader both bet on the same agentic AI platform within the same month, the burden of proof shifts to the skeptics. The question is no longer whether agentic AI fits into drug discovery workflows. It is how fast the rest of the industry follows.
Why Investors Bet on Agentic AI
The $13.5 million seed round crystallizes a thesis that has been hardening across the venture terrain for eighteen months: agentic AI, not generative chat, is the layer that finally makes software eat drug discovery.
Andreessen Horowitz partner Jorge Conde put it plainly: scientific AI will transform discovery the way cloud transformed software. The firm's own fund history traces the arc. A $200 million Bio + Health fund in 2015 grew to $450 million in 2017, $1.5 billion in 2022, and a $700 million vehicle in January 2026, part of a $15 billion raise across five new funds that month. The capital is there. The conviction is specific. a16z praised Phylo's alignment with scientist workflows rather than generic chat interfaces. Menlo Ventures echoed that structured lab integrations differentiate the company. Both funds had watched the previous generation, notebook-style platforms that captured data but lacked autonomous planning, saturate without delivering the productivity step-change LPs expect.
PitchBook counted $3.2 billion deployed across 135 AI drug development startups in the trailing twelve months. Nvidia's Inception program now tracks more than 5,000 healthcare and life sciences startups, digital health leading at 2,000-plus members. The raw deal flow is massive. The filter has tightened. Investors now seek systems that bridge experimental design, execution, and provenance in a single loop. Phylo's Biomni Lab, branded an Integrated Biology Environment, unites a large-language-model agent, a graphical interface, and more than 300 connectors to databases and instruments. The agent plans multi-step procedures, searches literature, generates code, and dispatches jobs to lab robots with minimal prompts. Safeguards include action logs and versioned runs so scientists can audit every step.
The friction is real. Enterprise adoption cycles in regulated biology remain lengthy. Biosecurity experts warn that agentic systems may lower barriers for malicious experimentation. The Bulletin of the Atomic Scientists recommends layered safeguards, including human-in-the-loop controls before executing sensitive protocols. Phylo states that Biomni denies suspicious requests and flags them for review, yet independent audits are pending. Scientific AI agents occasionally hallucinate protocols, leading to erroneous conclusions. Laboratories must validate outputs through orthogonal assays and peer review.
Investors are betting that the loop — design, execute, record, learn — compresses the timelines that have defined pharma R&D for decades. The Phylo round is a down payment on that bet.
Incumbents Scramble for Agentic Layer
The agentic AI wave that Phylo rode into pharma partnerships did not arrive in a vacuum. Established AI-native drug discovery companies, such as Insitro and Recursion Pharmaceuticals, have spent the last decade building proprietary data stacks, automated wet labs, and predictive models. But the April 2026 enterprise inflection point, when Microsoft, Google, Salesforce, and Anthropic all shipped production-grade agentic platforms within weeks
Working in frontier tech? Zero G Talent tracks the openings: see every open Insitro role, browse frontier tech jobs, openings at Phylo, and the people building the field.