The CSO and the Data-Science/Finance Chiefs
Formation Bio named Michael Ehlers chief scientific officer and head of R&D in July 2026, installing the executive who oversaw six approved medicines across five indications and four modalities at his prior post to run the company's AI-driven drug engine. The hire follows a sequence layering senior leaders who combine deep pharmaceutical operations with the data-science and finance discipline needed to compress clinical timelines.
The announcement positions Ehlers as the single executive overseeing portfolio strategy from translational science through clinical development while guiding the scientific direction of Formation Bio's AI platform. His track record — more than 40 programs advanced into the clinic — is cited as evidence he can address the bottleneck where AI-generated candidates meet the costly, slow reality of human trials. Benjamine Liu, Formation Bio's co-founder and CEO, framed the choice around the intersection of "the best scientists and the best technology," a formulation that recurs in the company's public messaging.
Ehlers is the most recent addition, but not the only one. In September 2025, Formation Bio disclosed four simultaneous appointments: Louis Brenner, MD, as chief medical officer; Daniel Neil, PhD, as chief technology officer; Frank D'Amelio as strategic advisor and co-chair of the Investment Advisory Committee; and Dashyant Dhanak, PhD, as senior advisor and committee member alongside Mikael Dolsten, MD, PhD. The release describes Neil's background as a "rare combination of deep machine learning and AI expertise, executive leadership, and biomedical industry experience": a profile matching the company's stated need to move beyond workflow automation toward augmented decision-making and, eventually, predictive toxicity and efficacy models.
D'Amelio brings nearly 15 years as Pfizer's chief financial officer and executive vice president of business operations and global supply, plus three decades of broader leadership. His quoted statement argues that Formation Bio's "NewCo model" creates a pathway for promising assets to accelerate through clinical value inflections while preserving safety and efficacy standards: a direct endorsement of the company's asset-acquisition strategy. Dolsten's inclusion, former president of worldwide R&D at Pfizer, signals access to large-pharma governance networks.
The board composition reinforces the financial and strategic orientation. As part of the $372 million Series D announced in June 2024, Scott Kupor of Andreessen Horowitz and Alfred Lin of Sequoia joined alongside existing director Michael Moritz of Sequoia Heritage and observer Kareem Zaki of Thrive Capital. That capital was earmarked for acquiring and in-licensing candidate drugs and expanding AI capabilities.
Ehlers' own statement captures the internal logic: "I have spent years watching good science get directed toward the wrong patients, or stuck in development timelines that have not meaningfully improved in decades." The leadership slate assembled around him — spanning R&D, clinical, technology, finance, and board governance — is structured to attack both problems simultaneously.
Platform Acceleration: From Asset to Clinic
The September 2025 leadership expansion brought in executives whose mandates map onto the two levers determining how fast an AI-native pharma turns external assets into clinical candidates: evaluation throughput and development velocity. The company's disclosures frame the hires as embedding technology deeper into every stage (from asset identification through trial execution), and the board's live hiring data confirms the build-out is still running.
Current open roles — a Senior Director or Vice President of Product, a Director of Product for Applied Data Science, a Director of Data Science, and a Director of Business Development focused on Search & Evaluation — each sit at a different choke point in the acquisition-to-clinic pipeline. The Business Development lead owns the top of the funnel: sourcing, triaging, and running diligence on in-licensing opportunities. The product and data-science leads own the engine that evaluates those opportunities: predictive models for toxicity, tolerability, and efficacy that the company says it is training to "steer high quality R&D decisions." When the same organization controls both deal flow and the analytical engine that scores it, the feedback loop compresses. A candidate that once required months of manual scientific review can be scored, ranked, and either advanced or killed in days.
Formation Bio's platform already demonstrates this compression in clinical operations. The AI system Muse, launched in November 2024 with OpenAI and Sanofi, reduced patient recruitment campaign creation from two months to minutes. Adverse-event report generation dropped from hours to minutes. The stated roadmap extends that pattern upstream: short-term workflow automation, medium-term "AI R&D Scientist" decision support, long-term predictive models for toxicity and efficacy. The product and data-science roles now open are tasked with delivering the medium-term layer: the augmented decision-making between raw asset data and a go/no-go investment call.
The asset acquisition target is explicit: 10 to 15 programs over the next three to five years, up from three clinical candidates today (a Phase 3 chronic hand eczema asset, an atopic dermatitis/notalgia paresthetica program, and a knee osteoarthritis candidate). At that pace, the evaluation engine must process hundreds of opportunities per year. The board's live listings show Formation Bio actively recruiting for an Associate Director of Data Science specializing in Real World Data: a signal that the diligence stack is incorporating external evidence streams (claims, registries, EHR) to de-risk assets before commitment. That capability, paired with a dedicated Search & Evaluation lead, creates a flywheel: better data ingestion feeds better models, which sharpen acquisition criteria, which yield higher-quality assets that generate richer clinical data for the next model iteration.
Capital follows the same logic. The Series D was explicitly earmarked to "grow drug pipeline and continue expanding AI-driven drug development platform." The valuation step-up reflected investor confidence that the platform could deploy capital efficiently: each acquired asset would move faster and cheaper through the clinic than industry benchmarks. The leadership hires operationalize that thesis. If the product and data-science teams deliver on the medium-term roadmap, the cost per successful program drops, portfolio throughput rises, and the 10–15 asset target becomes credible rather than aspirational.
The bottleneck Formation Bio is trying to break is structural: discovery output has doubled over the past decade while approvals have flatlined around 60 per year, because clinical development remains too slow and expensive to absorb the surplus. The company's bet is that AI, embedded end-to-end and owned by the same team that picks the assets, can widen that bottleneck. The September 2025 hires (and the product, data-science, and search roles still open) are the mechanism.
Investors Bet on the Model
The $372 million Series D closed in June 2024 sent the clearest signal yet that sophisticated capital views Formation Bio's model as investable at scale. Andreessen Horowitz led. Sanofi and Thrive Capital participated. The deal valued the company at $1.7 billion, per PitchBook data cited by CNBC in June 2025.
Scott Kupor, managing partner at a16z, framed the bet in industry terms: "Pharma represents one of the biggest industries in the world, and there is immense potential to make the drug development process more efficient." Alfred Lin of Sequoia went further: "We are in the early stages of AI driving significant efficiencies in pharma and biotech, and Formation Bio is well-positioned to lead this transformation." Michael Moritz, senior advisor to Sequoia Heritage, described the coming "avalanche of promising looking drugs that merit enquiry" and positioned Formation as the filter: "Formation Bio has the technology and know-how to help the industry sort the wheat from the chaff." Paul Hudson, Sanofi's chief executive, put it plainly: "At Sanofi, we're all in on AI. And we are proud to partner with and invest in Formation Bio."
Strategic capital matters more than headline numbers. Sanofi's investment came with a product partnership: the Muse AI system, co-developed with OpenAI, now runs Phase 3 multiple sclerosis trial recruitment for the French pharma giant. That deal converts investor conviction into recurring platform revenue and de-risks the thesis that Formation's tools work on assets it doesn't own.
The clearest proof of model viability arrived in September 2025. Formation announced an out-licensing transaction valued at up to €545 million in upfront, development, and commercial milestones plus royalties, according to Formation Bio's September 2025 press release. The company did not name the counterparty or the asset. The structure — milestone-heavy, royalty-bearing — mirrors the "NewCo" model Frank D'Amelio, the former Pfizer CFO who joined Formation's leadership team in 2025, has championed. D'Amelio argued the traditional pharmaceutical model "simply can't keep pace with today's rate of scientific discovery" and called Formation's approach "a lift-all-boats model for the industry: unlocking the potential of promising drug candidates that need the right partner, capital, and capabilities to reach patients." His presence upgrades the financial discipline around each asset vehicle. Investors tracking Formation now have a CFO who has guided a $100 billion revenue company through patent cliffs and megadeals.
The €545 million deal also validates the acquisition thesis. Formation's playbook: buy stalled clinical assets cheap, accelerate them with AI, then sell or partner after Phase 2 or 3 data. Each successful out-license compounds the track record and lowers the cost of capital for the next acquisition.
Sector consolidation reinforces the M&A optionality. Recursion Pharmaceuticals absorbed Exscientia in a stock deal that closed in 2024, creating a combined entity with a larger data stack and broader pipeline. Insitro acquired CombinAbleAI in 2024 to expand its generative chemistry capabilities. BenevolentAI restructured multiple times and pivoted back to its "TechBio roots" with a new CEO and a partnership with Merck KGaA. Atomwise, once a high-flier, pared down and installed a biotech veteran as chief executive while securing a five-drug deal with Sanofi and an evaluation agreement with Pfizer. In this environment, Formation's balance sheet ($600 million raised since 2016) and its NewCo structure make it both a buyer of distressed assets and a potential takeout target for a large pharma seeking an AI-native development engine. Sanofi already owns a stake and uses the platform. The logic of deeper integration is visible.
Analyst coverage remains thin; Formation is private. But the investor syndicate — a16z, Sequoia, Thrive, Sanofi — functions as a de facto valuation committee. Each follow-on investment or partnership re-prices the option. The next inflection will come when the first Formation-advanced asset files for regulatory approval. Until then, the €545 million milestone deal and the Sanofi-MS partnership are the only external price discoveries. They suggest the market is starting to believe the compression story.
The Arms Race in Data and Talent
Recursion Pharmaceuticals has spent more than a decade building what its leadership calls one of the world's largest proprietary biological datasets: over 50 petabytes spanning phenomics, transcriptomics, proteomics, ADME, and de-identified patient data, with an automated wet lab capturing millions of cell experiments per week. That infrastructure didn't appear overnight. When Recursion started, its CEO Chris Gibson has said, roughly ten companies called themselves tech-enabled drug discovery players; today there are hundreds, "probably too many." The Exscientia absorption announced in 2024 (a stock deal valuing the combined entity around $688 million at signing) reads as a direct answer to that crowding. By folding Exscientia's precision chemistry and automation stack into Recursion's phenomics engine, the merged company gains a more complete loop from target identification through lead optimization, exactly the kind of end-to-end platform Formation Bio is now assembling with its own hires.
The data competition is visible in what Recursion chooses to release. In August 2023 the company open-sourced 2.2 million images of human cells across nearly the entire genome, calling it the largest such release in the space. That move serves two masters: it seeds the community with training data that improves the field's models (benefiting Recursion's own downstream inference) and it signals scale to partners and investors. The Genentech collaboration, described by Recursion as "the largest collaboration tech-enabled drug discovery... one of the largest blue sky collaborations to go for a decade across the whole of Neuroscience," followed similar logic. Genentech brings clinical development muscle; Recursion brings the map. Formation Bio's OpenAI-Sanofi partnership, announced alongside its Series D in June 2024, mirrors that structure: a big pharma partner, an AI lab, and a tech-native biotech in the middle.
Recursion's clinical funnel is also tightening. The company initiated five trials in 2022 and planned a sixth by mid-2023; its Phase 1 MALT1 inhibitor REC-3565 dosed its first patient in 2024. That narrative (efficiency gains from platform maturity) is exactly the story Formation Bio tells with its own trial-speed metrics: faster study startup, recruitment, data management, and database lock versus industry benchmarks. When two competitors cite the same efficiency levers, the market starts treating platform velocity as a baseline expectation rather than a differentiator.
Elsewhere, the consolidation logic repeats, with Insitro, BenevolentAI, and Atomwise making similar moves as previously noted. The French pharma giant is simultaneously betting on external AI partners: Formation Bio, Recursion, Atomwise, and the Illumina Billion Cell Atlas, which added Formation Bio as a member in July 2026.
The talent market reflects this convergence. Formation Bio's September 2025 leadership slate (Daniel Neil as CTO from Tessera (ex-BenevolentAI), plus hires from Moderna, Google DeepMind, and Pfizer) reads like a draft board assembled from the same pool Recursion, Insitro, and BenevolentAI fish. Neil scaled BenevolentAI's technology team to more than 150 people through growth, restructuring, and a public listing; that playbook now applies at Formation Bio, which plans 10–15 assets in three to five years.
The competitive dynamic is no longer about who has the best model. It is about who can integrate proprietary data, automated wet lab throughput, clinical operations, and capital allocation into a repeatable asset factory. Formation Bio's CSO hire and product-data-science promotions bet that the factory's control room needs a single leader who speaks both biology and platform. Recursion's Exscientia deal and Genentech pact bet that the factory needs a chemistry wing and a pharma distribution channel. Insitro's CombinAbleAI buy and BenevolentAI's Merck link bet that the factory needs generative chemistry and a big-pharma co-pilot. The winners will be the ones who stop describing their platforms and start shipping approved drugs; the hiring sprees across all four companies suggest they know it.
Two Cities, One Talent Pool
Formation Bio's headquarters has been in New York City since its 2016 launch as TrialSpark, but the company's recent hiring pattern reveals a deliberate dual-hub strategy within the Northeast corridor. First-party board data from Zero G Talent shows that six of the company's open senior roles (spanning product, data science, business development, and general management) list both Boston and New York as interchangeable locations.
| Role | Location | Salary Band |
|---|---|---|
| General Manager (Drug Asset Lead) | Boston & New York | $311,000–$388,500 |
| Senior Director / Vice President of Product | Boston & New York | $270,000–$353,000 |
| Director of Product for Applied Data Science | Boston & New York | $245,000–$307,000 |
| Director of Data Science | Boston & New York | $235,000–$307,000 |
| Director of Business Development, Search & Evaluation | Boston only | $235,000–$307,000 |
| Associate Director of Data Science, Real World Data | Boston & New York | $213,500–$267,000 |
| Overall (22 salaried roles) | Boston & New York | $163,000–$307,000 (median $232,000) |
The board's overall salary band for Formation Bio's 22 salaried roles aligns with the premium the corridor commands for hybrid AI-bio talent.
This dual-location posture is not cosmetic. The NY-Boston corridor concentrates the three inputs Formation Bio's model demands: late-stage clinical operations expertise from Big Pharma's New Jersey and New York offices, computational biology and machine-learning talent from Boston's academic labs and the Kendall Square ecosystem, and capital from the venture firms that have backed the company's $600 million in funding and $1.7 billion valuation (PitchBook, as of June 2025). Benjamine Liu, co-founder and CEO, has said the company was built around that idea: a formulation that only works if the organization sits where both populations already live. The board data shows Formation Bio posting two new roles in the past seven days alone, both with the same Boston-New York flexibility, signaling that the corridor strategy is accelerating, not settling.
The talent implications extend beyond Formation Bio. Recursion Pharmaceuticals, Insitro, and the newly merged Recursion-Exscientia entity all maintain significant Boston footprints while keeping executive or commercial teams in New York or San Francisco. The corridor's density also supports the "NewCo" model Formation Bio uses to spin out asset-specific vehicles: legal, regulatory, and CMC consultants who know the FDA's Northeast corridor reviewers are concentrated within a four-hour train ride.
What the hiring map makes clear is that Formation Bio treats geography as a platform variable. By posting every senior data-science and product role as portable between Boston and New York, the company expands its candidate pool compared to a single-city search: a practical advantage when the target profile requires fluency in both clinical trial operations and large-language-model fine-tuning. The corridor's transit link (Acela, three hours forty minutes door-to-door) makes weekly in-person collaboration feasible, which matters for a company that ships AI-generated adverse-event reports in minutes and runs study startup faster than industry benchmarks. Other AI-native pharmas will likely follow the same pattern: the next wave of chief scientific officers and heads of product will negotiate dual-city clauses as standard, and the corridor's salary bands will continue to lead the sector.
The Factory Floor
Ehlers has previously described this bottleneck. The factory he now helps run — spanning two cities, a $1.7 billion valuation, and a growing roster of rivals racing to build their own — will be measured by whether it finally breaks that bottleneck. The first Formation-advanced asset filing for approval will tell the market if the compression story holds.
Working in frontier tech? Zero G Talent tracks the openings: see every open Formation Bio role, browse frontier tech jobs, the companies hiring, and the people building the field.