Andrew White

Andrew White is an American chemical engineer. He is the co-founder and head of science of FutureHouse, the San Francisco AI-for-science nonprofit founded in September 2023 with Sam Rodriques and primarily funded by Eric Schmidt, and an associate professor of chemical engineering at the University of Rochester on academic leave. He is the lead developer of the chemistry-language-model agent line that includes ChemCrow, PaperQA, WikiCrow, and the ether0 chemistry-reasoning model.

At a glance

• Education: BS in chemical engineering, Rose-Hulman Institute of Technology (2008); PhD in chemical engineering, University of Washington (2013), advised by Shaoyi Jiang.
• Current role: Co-founder and head of science of FutureHouse since September 2023; co-founder and chief technology officer of Edison Scientific since November 2025; associate professor of chemical engineering at the University of Rochester, on academic leave.
• Key contributions: Senior author of ChemCrow (Nature Machine Intelligence, 2024) and PaperQA (2023); lead developer of WikiCrow (autonomous Wikipedia-style article generation); co-developer of ether0 (June 2025); GPT-4 red-team participant for OpenAI in 2022.
• Recognition: NSF CAREER Award (2018); NIH Outstanding Investigator Award (R35, MIRA) (2020); AIChE COMSEF Young Investigator Award (2020); Outstanding Early Career Researcher Award, Royal Society of Chemistry's Digital Discovery (2024).
• X / Twitter: @andrewwhite01
• LinkedIn: andrewdwhite
• Personal site / lab: thewhitelab.org
• Google Scholar: Andrew D. White
• GitHub: whitead

Origins

White is American. He completed undergraduate study at the Rose-Hulman Institute of Technology in Terre Haute, Indiana, from 2004 through 2008, earning a bachelor of science in chemical engineering, with a year abroad in Magdeburg, Germany, at the Otto-von-Guericke Universität and the Max Planck Institute for Dynamics of Complex Technical Systems.

After Rose-Hulman, White entered the chemical-engineering doctoral program at the University of Washington in Seattle, completing a PhD in 2013 under Shaoyi Jiang. The doctoral research focused on non-fouling biomimetic surfaces with computational modeling, with particular attention to materials engineered to resist the foreign-body response in applications including brain-machine interfaces. The thesis combined molecular-dynamics simulation with experimental surface chemistry.

From 2013 through 2014, White was a postdoctoral fellow at the University of Chicago Institute for Biophysical Dynamics in the laboratory of Gregory Voth. The Chicago postdoctoral period developed methods for combining first-principles molecular simulation with experimental data.

Career

In 2015 White joined the University of Rochester as assistant professor of chemical engineering, with joint appointments in chemistry, biophysics, materials science, and the data-science program. The Rochester laboratory initially continued the molecular-simulation and biomaterials research lines from the doctoral period, with a teaching portfolio that included the elective CHE 478 (Machine Learning of Molecules and Materials). White was promoted to associate professor and the laboratory expanded into machine-learning methods for chemistry across the late 2010s.

A 2019 sabbatical at the Institute for Pure and Applied Mathematics (IPAM) at UCLA was a pivot point. The IPAM program brought him into contact with Yann LeCun and other deep-learning researchers and oriented his subsequent research toward language-model applications in chemistry. Following the sabbatical White published the open-access textbook Deep Learning for Molecules and Materials.

In 2022 White was one of approximately fifty external experts that OpenAI recruited as part of the GPT-4 red-team program, with a remit to test the model for potential misuse in chemistry contexts. The red-teaming experience produced subsequent commentary on AI-chemistry safety and biosecurity, including Financial Times coverage.

In April 2023 White and collaborators including Andres Bran and Philippe Schwaller at EPFL released the ChemCrow preprint, characterized by the authors as the first large-language-model agent for chemistry. The paper was published in Nature Machine Intelligence in May 2024.

In September 2023 White co-founded FutureHouse with Sam Rodriques, then a group leader at the Francis Crick Institute in London. The pair connected in late 2022 around shared interest in foundation models for scientific research, with Rodriques contributing the bioengineering record from MIT and Crick and White contributing the chemistry-LLM agent line. Rodriques took the chief-executive role and White became head of science. White concurrently took academic leave from Rochester.

In November 2025 FutureHouse spun out Edison Scientific as the commercial vehicle for the AI-scientist platform, with White as co-founder and chief technology officer. Edison launched the Kosmos AI-scientist agent at spinout and raised a $70 million seed round in December 2025 at a $250 million valuation, co-led by Spark Capital and Triatomic Capital.

Affiliations

• University of Washington Department of Chemical Engineering: PhD researcher in the Jiang group, 2009 to 2013.
• University of Chicago Institute for Biophysical Dynamics: Postdoctoral fellow in the Voth group, 2013 to 2014.
• University of Rochester: Assistant and then associate professor of chemical engineering, 2015 to present, with joint appointments in chemistry, biophysics, materials science, and data science. On academic leave from 2023.
• FutureHouse: Co-founder and head of science, September 2023 to present.
• Edison Scientific: Co-founder and chief technology officer, November 2025 to present.

Notable contributions

The Google Scholar record reports approximately 7,800 cumulative citations and an h-index of 41 across more than seventy-five peer-reviewed publications.

• Promoting transparency and reproducibility in enhanced molecular simulations (Nature Methods, 2019). Co-author on the standards paper for enhanced-sampling reproducibility.
• Assessment of chemistry knowledge in large language models that generate code (Digital Discovery, 2023). Lead author on the early chemistry-LLM-evaluation paper introducing an open-source benchmark for code-generating language models.
• ChemCrow: Augmenting large-language models with chemistry tools (Nature Machine Intelligence, 2024). Senior author with Andres Bran, Sam Cox, Oliver Schilter, Carlo Baldassari, and Philippe Schwaller. ChemCrow integrated 18 chemistry-domain tools and demonstrated autonomous synthesis planning.
• PaperQA (December 2023). Senior author with Jakub Lála, Odhran O'Donoghue, Aleksandar Shtedritski, Sam Cox, and Sam Rodriques on the retrieval-augmented generative-agent paper that introduced FutureHouse's first public AI-for-science system. PaperQA matched expert human researchers on the LitQA benchmark.
• WikiCrow (2023 to 2024). Lead developer of the autonomous Wikipedia-style article generation system that produced cited articles for human protein-coding genes lacking Wikipedia coverage. PhD and postdoctoral biology researchers rated the WikiCrow articles as more accurate on average than human-written Wikipedia entries on the same topics.
• Language agents achieve superhuman synthesis of scientific knowledge (September 2024). Co-author on the FutureHouse paper documenting that language-agent systems exceed expert humans on scientific-literature-synthesis tasks.
• ether0 (June 2025). Co-developer of the 24-billion-parameter open-weights chemistry-reasoning model, characterized as the first reinforcement-learning-trained reasoning model in a scientific domain. The accompanying paper Training a Scientific Reasoning Model for Chemistry was covered by Nature at the launch.
• Deep Learning for Molecules and Materials (2021 to present). Open-access self-published textbook on deep learning for chemistry and materials.
• Public-talk record. Language is the Future of Chemistry at Applied Machine Learning Days (2023); Automating Scientific Discovery on The Cognitive Revolution podcast (December 2024); Building an AI Scientist to Automate Discovery on the Existential Hope podcast (2024).

Investments and boards

The entries below are limited to AI, semiconductors, datacenters, software, and energy.

• FutureHouse (AI): Co-founder and head of science, September 2023 to present. San Francisco-based 501(c)(3) nonprofit AI-for-science research laboratory. Primarily funded by Eric Schmidt; cumulative funding not publicly disclosed.
• Edison Scientific (AI): Co-founder and chief technology officer, November 2025 to present. FutureHouse commercial spinout building the Kosmos AI-scientist platform. Raised a $70 million seed round in December 2025 at a $250 million valuation, co-led by Spark Capital and Triatomic Capital.
• Acellera Therapeutics (AI / Bio): Advisory board member, June 2024 to present. Computational drug-discovery platform building the QuantumBind and ACEGEN AI-driven discovery products.

No other public investor activity on record in AI, semiconductors, datacenters, software, or energy as of May 2026.

Network

White's longest-running professional collaboration is with Sam Rodriques, his FutureHouse and Edison Scientific co-founder. The pair connected in late 2022 around shared interest in language-model agents for scientific research, with White contributing the chemistry-LLM agent line and Rodriques contributing the bioengineering record from MIT and the Francis Crick Institute. FutureHouse research staff include Jakub Lála, Odhran O'Donoghue, Aleksandar Shtedritski, and Sam Cox as PaperQA co-authors and Edison Scientific founding researchers. Sam Cox previously worked in the Rochester laboratory and bridges the academic and FutureHouse periods.

The ChemCrow collaboration produced an ongoing research relationship with Philippe Schwaller and Andres Bran at EPFL. The doctoral relationship with Shaoyi Jiang at the University of Washington and the postdoctoral relationship with Gregory Voth at the University of Chicago anchor the biomaterials-and-simulation network.

The Eric Schmidt and Schmidt Sciences funding relationship is the principal financial anchor for the FutureHouse operation. Among AI-for-science peer founders, his position runs in parallel with Demis Hassabis at Isomorphic Labs, the leadership at EvolutionaryScale, Inceptive, Lila Sciences, and Vik Bajaj at Project Prometheus.

Position in the field

As of May 2026, White occupies a distinctive position among AI-for-science research leaders through the combination of the academic-research record at the University of Rochester, the GPT-4 red-team participation, the ChemCrow chemistry-LLM-agent authorship, and the FutureHouse and Edison Scientific co-founding role.

Industry coverage has consistently characterized FutureHouse under the Rodriques-and-White co-leadership as one of the principal nonprofit AI-for-science research organizations globally, alongside Isomorphic Labs, EvolutionaryScale, Inceptive, and Lila Sciences. The chemistry-LLM-agent line that includes ChemCrow, PaperQA, WikiCrow, and ether0 is among the most-cited research lines in AI-for-chemistry coverage of the 2023 to 2026 period.

The chemical-engineer-to-AI-research-leader arc is unusual among AI-for-science executive teams. Most peer chief executives in AI-for-science laboratories have come from drug-discovery or computational-biology backgrounds, with relatively few from chemical engineering specifically. The Deep Learning for Molecules and Materials textbook and the @andrewwhite01 social-media reach have placed White as a public-facing figure for the AI-for-chemistry research community.

Outlook

Open questions over the next 6 to 18 months:

• Edison Scientific funding cadence. Whether Edison Scientific raises a Series A in 2026, the lead investor and valuation, and the capital-deployment plan for the Kosmos and chemistry-agent platforms.
• Chemistry-reasoning-model trajectory. The cadence of follow-on chemistry-reasoning-model releases beyond ether0, and the relative positioning against general-purpose frontier reasoning models on chemistry-specific benchmarks.
• University of Rochester return decision. Whether White returns to the Rochester faculty position at the conclusion of the academic leave or transitions fully to the FutureHouse and Edison Scientific roles.
• FutureHouse research output. Continued cadence of autonomous-research-agent releases from the nonprofit relative to the Edison Scientific commercial operation.
• Schmidt-affiliated funding trajectory. Continued personal commitment from Eric Schmidt and Schmidt Sciences participation as FutureHouse expands operating scope.

Sources

• Andrew White at FutureHouse. FutureHouse team-page biography of the head of science and co-founder.
• Andrew White personal site / The White Lab. Personal site documenting research interests, the FutureHouse and Edison Scientific co-founding, and the academic-leave status.
• Andrew D. White on the University of Rochester chemical engineering faculty page. University of Rochester faculty profile documenting the associate professor title, education, and selected honors.
• Andrew D. White on Google Scholar. Google Scholar record with citation counts and the principal publication record.
• Andrew White on LinkedIn. LinkedIn profile documenting the FutureHouse, Edison Scientific, Rochester, Chicago, and Washington chronology.
• Rose-Hulman Alumnus at the Intersection of Artificial Intelligence and Chemistry. Rose-Hulman Institute of Technology 2024 alumni profile documenting the 2008 BS and the FutureHouse co-founding.
• Faculty Member is Helping Former Google CEO Reshape Future of Scientific Research. University of Rochester Department of Chemical Engineering announcement of White's FutureHouse co-founding and academic leave.
• PaperQA: Retrieval-Augmented Generative Agent for Scientific Research. December 2023 arXiv preprint introducing the FutureHouse PaperQA system, with White as senior author.
• ChemCrow: Augmenting large-language models with chemistry tools. Nature Machine Intelligence May 2024 paper introducing the ChemCrow chemistry-LLM agent, with White as senior author.
• WikiCrow: PaperQA2 Superhuman scientific literature search. FutureHouse research announcement for the PaperQA2 and WikiCrow systems.
• ether0: a scientific reasoning model for chemistry. FutureHouse June 2025 announcement of the ether0 chemistry-reasoning model.
• Training a Scientific Reasoning Model for Chemistry. June 2025 arXiv preprint describing the ether0 model training methodology, with White as co-author.
• Exclusive: Start-up FutureHouse debuts powerful AI 'reasoning model' for science. Nature June 2025 coverage of the ether0 launch.
• Andrew White Joins Acellera Therapeutics' Advisory Board. Acellera Therapeutics June 2024 announcement of White's advisory-board appointment.
• Andrew White: Building an AI Scientist to Automate Discovery. Existential Hope 2024 podcast covering the IPAM sabbatical, the chemistry-LLM-agent line, and the FutureHouse founding.
• Accelerating scientific discovery with AI. MIT News June 2025 profile of FutureHouse, with details on the Rodriques and White founding partnership.
• Photo: Andrew White at FutureHouse, FutureHouse team page.