Eric Ho

Eric Ho is an American entrepreneur and the co-founder and chief executive officer of Goodfire, the San Francisco mechanistic-interpretability company he established in June 2024 with Daniel Balsam and Tom McGrath. The company develops Ember, a hosted interpretability platform that decodes the internal computations of large neural networks and exposes them as programmable features. As of May 2026, Ho leads Goodfire after a $150 million Series B in February 2026 at a $1.25 billion valuation, the third financing in eighteen months following an August 2024 seed and an April 2025 Series A in which Anthropic made its first equity investment in another startup.

At a glance

• Education: Bachelor's degree in computer science, Yale University, 2016 (graduated as Eric Ho '16).
• Current role: Co-founder and chief executive officer of Goodfire, since June 2024.
• Key contributions: Co-founder of RippleMatch (2016) and its president and chief technology officer through December 2023; co-founder and chief executive of Goodfire (2024); architect of the founding team that combined an operating chief executive with the academic-research credentials of co-founders Daniel Balsam (chief technology officer) and Tom McGrath (chief scientist, formerly of Google DeepMind); recruited Chris Olah of Anthropic as advisor.
• Public profile: Sequoia Capital Training Data podcast appearance in July 2025; Forbes 30 Under 30 for Enterprise Technology in 2022; LessWrong essay Some for-profit AI alignment org ideas in December 2023 that articulated his pivot from recruiting software to AI safety.
• X / Twitter: @ericho_goodfire
• LinkedIn: eric-ho-53981862
• LessWrong: eric-ho

Origins

Ho is American and attended Yale University in New Haven, Connecticut, graduating with a bachelor's degree in computer science in 2016. The Yale Daily News profile of RippleMatch in February 2019 lists him as Eric Ho '16 and describes the founding of the company in a Yale dorm room with co-founder Andrew Myers '18 in 2016. According to public profiles, Ho held a software-engineering internship at Facebook during the undergraduate period.

The recurring account in industry coverage and in his own LinkedIn announcement of his RippleMatch departure is that he founded the company out of a personal frustration with the early-career hiring process during his Yale senior year. The framing positions the original company motivation as one of access and efficiency in entry-level recruiting, distinct from the later AI-safety motivation that drove the Goodfire founding.

Career

After graduating from Yale in 2016, Ho co-founded RippleMatch in June 2016 with Andrew Myers and Troy LeCaire. The company was a New York-headquartered AI-driven early-career recruiting platform; Ho took the title of president and chief technology officer, with Myers as chief executive. RippleMatch raised approximately $79.9 million across multiple rounds, with the principal round a $45 million Series B led by the Growth Equity business at Goldman Sachs Asset Management in June 2022. By the time of his departure RippleMatch counted employers including SAP, General Mills, eBay, EY, and AB InBev among its customers. In 2022 Ho was named to the Forbes 30 Under 30 list for Enterprise Technology alongside Myers.

The strategic shift began in 2023. In December 2023 Ho announced his transition from operator to board member and advisor at RippleMatch in a LinkedIn post that framed the next chapter as work on AI safety. The same month he published Some for-profit AI alignment org ideas on LessWrong, arguing that for-profit organizations could scale AI safety research faster than non-profit alternatives, with proposals for red-teaming services and constrained-agent developer frameworks among the specific commercial directions.

Goodfire was incorporated in June 2024 with Ho as chief executive officer, Daniel Balsam (formerly the founding engineer at RippleMatch) as chief technology officer, and Tom McGrath (formerly senior research scientist and founder of the mechanistic-interpretability team at Google DeepMind) as chief scientist. The founding thesis was that mechanistic interpretability could be turned into a commercial layer of the AI stack rather than remaining an academic-research activity confined to frontier laboratories. McGrath's prior leadership of the DeepMind interpretability team supplied the scientific anchor; Ho and Balsam supplied the operating and engineering anchor.

The fundraising trajectory has been compressed. The August 2024 seed round of $7 million was led by Lightspeed Venture Partners with participation from Menlo Ventures, South Park Commons, Work-Bench, Juniper Ventures, and angel investors; Lightspeed partner Nnamdi Iregbulem led the round. The April 2025 $50 million Series A at a $250 million valuation was led by Menlo Ventures with Anthropic participating, in what coverage including Tech Startups reported as Anthropic's first equity investment in another startup. Menlo partner Deedy Das led the Series A and joined the Goodfire board. The Series A coincided with the public launch of the Ember platform that Goodfire had previewed in December 2024.

In February 2026 Goodfire raised a $150 million Series B at a $1.25 billion valuation, led by B Capital with participation from Juniper Ventures, DFJ Growth, Salesforce Ventures, Menlo Ventures, Lightspeed, South Park Commons, Wing Venture Capital, and Eric Schmidt. Cumulative funding through May 2026 is approximately $209 million.

Affiliations

• RippleMatch: Co-founder, President, and Chief Technology Officer, June 2016 to December 2023; Board member and advisor, December 2023 to present.
• Goodfire: Co-founder and Chief Executive Officer, June 2024 to present.

Notable contributions

Ho's published output sits at the founding-and-operating end of the spectrum rather than the published-research end. His record is concentrated on company-formation, fundraising, public communication, and the recruitment of senior research-anchor co-founders rather than on individual research authorship.

• RippleMatch co-founding (2016). Co-founded the AI-driven early-career recruiting platform with Andrew Myers and Troy LeCaire while a Yale senior. Held the president and chief technology officer titles for seven and a half years, through approximately $79.9 million in cumulative venture capital.
• Forbes 30 Under 30 (2022). Named to the Forbes Enterprise Technology list alongside Andrew Myers for the RippleMatch work.
• Some for-profit AI alignment org ideas (LessWrong, December 2023). Essay outlining the case for commercially-funded AI safety organizations, with specific proposals for red-teaming services, constrained agent frameworks, and a public-benefit-corporation governance structure. Published the same month Ho announced his RippleMatch departure.
• Goodfire co-founding (June 2024). Established the mechanistic-interpretability company with Daniel Balsam as chief technology officer and Tom McGrath as chief scientist; an unusual founding-team structure that paired an operating chief executive with engineering and academic-research credentials in the co-founders.
• Ember (December 2024 research preview, April 2025 platform launch). Commercial mechanistic-interpretability platform exposing sparse autoencoder features as a hosted application programming interface and software development kit. Initial public support for Llama 3.3 70B and Llama 3.1 8B; subsequent expansion to additional model families and customer-specific models.
• Public-talk record. Sequoia Capital Training Data podcast appearance in July 2025 hosted by Sonya Huang and Roelof Botha, in which Ho predicted that neural networks will be substantially decoded by 2028; recurring LinkedIn and X commentary on Goodfire's research output. Conference and long-form-interview appearances are limited relative to peer figures at frontier laboratories.
• Quoted positioning. Ho has been quoted in primary press coverage including the Series A announcement ("Nobody understands the mechanisms by which AI models fail, so no one knows how to fix them") and the Series B announcement ("Interpretability, for us, is the toolset for a new domain of science: a way to form hypotheses, run experiments, and ultimately design intelligence rather than stumbling into it").

Investments and boards

• RippleMatch (Software): Co-founder + Board member, 2016 to present (operating role through December 2023, board and advisor role since). Privately held early-career recruiting platform; approximately $79.9 million in cumulative funding through 2026, including the $45 million Series B in June 2022 led by Goldman Sachs Asset Management.
• Goodfire (AI): Co-founder and Chief Executive Officer, June 2024 to present. Privately held mechanistic-interpretability company; approximately $209 million in cumulative funding through three rounds, with the $150 million Series B in February 2026 at a $1.25 billion valuation.

No public personal angel-investor activity on record outside the RippleMatch and Goodfire founder roles in AI, semiconductors, datacenters, software, or energy as of May 2026.

Network

Ho's longest-running professional relationship is with Andrew Myers, his Yale-era co-founder at RippleMatch and the chief executive of the recruiting platform from 2016 onward. Daniel Balsam, his Goodfire chief technology officer co-founder, was the founding engineer at RippleMatch and joined Ho on the AI-safety pivot after a 2023 research sabbatical, making the working relationship continuous across both companies. Tom McGrath, the third Goodfire co-founder and chief scientist, joined the founding team after leaving Google DeepMind in March 2024 and a brief period at South Park Commons before connecting with Ho and Balsam.

The most distinctive external relationship is the advisory role of Chris Olah, the Anthropic co-founder who leads the company's interpretability research program and is widely cited as one of the pioneers of the mechanistic-interpretability research field. Olah's advisor relationship with Goodfire is unusual given his concurrent senior position at Anthropic, and reflects the alignment between Anthropic's research priorities and Goodfire's commercial positioning. The Anthropic Series A investment, the first equity investment in another startup that Anthropic has disclosed, deepened the institutional relationship beyond the personal-advisor link.

The investor network spans multiple venture firms and strategic backers. Lightspeed partner Nnamdi Iregbulem led the seed round; Menlo partner Deedy Das led the Series A and holds a Goodfire board seat; B Capital led the Series B. Strategic investors include Salesforce Ventures and Eric Schmidt. Through the Goodfire customer base, Ho has working relationships with Patrick Hsu, the co-founder of the Arc Institute and the Evo 2 DNA foundation-model team, who is cited in the Series A announcement as an early Ember collaborator on the application of mechanistic interpretability to the Arc Institute's biology research.

Position in the field

As of May 2026, Ho is the chief executive of the only commercial mechanistic-interpretability laboratory to have crossed the unicorn-valuation threshold. The Series B at a $1.25 billion valuation in February 2026 places Goodfire at a structurally distinctive position in the AI safety and governance market, separate both from the in-house interpretability programs at frontier laboratories such as Anthropic and Google DeepMind and from the smaller commercial-interpretability and governance startups in the same product category.

The career trajectory is unusual relative to peer chief executives in mechanistic-interpretability research. Ho's prior operating experience is in early-career recruiting software rather than in academic AI research, and his transition into AI safety was articulated through the December 2023 LessWrong essay rather than through a published-research record. Industry coverage including the Sequoia Capital Training Data podcast episode characterizes Goodfire's founding-team structure as a deliberate combination of an operating chief executive with the academic-research credentials concentrated in McGrath as chief scientist and Olah as advisor, rather than as a research-led company with the chief executive as the principal scientific voice.

The Anthropic Series A participation has been the most-cited external validation of the positioning. Anthropic CEO Dario Amodei was quoted in the Series A announcement framing the investment as a reflection of Anthropic's confidence in mechanistic interpretability "as a critical foundation for the responsible development of powerful AI." The relationship between Anthropic's in-house interpretability program and Goodfire's commercial offering is one of cooperation rather than head-on competition, supported by the Olah advisor role and the Anthropic strategic investment.

Outlook

Open questions over the next 6 to 18 months:

• Ember adoption. Whether the platform's customer base scales materially across AI laboratories, regulated-industry enterprises, and biology-and-genomics research groups, and whether the Arc Institute Evo 2 collaboration produces additional published case studies of mechanistic-interpretability application beyond language-model contexts.
• Strategic-investor depth. Whether the Anthropic relationship deepens beyond capital participation into formal product collaborations, and whether the Salesforce Ventures Series B participation produces enterprise-distribution channels for Ember.
• Public-communication cadence. Whether Ho's interview record expands beyond the Sequoia Training Data appearance into more frequent conference keynotes, peer-reviewed publications, or long-form AI-safety interviews.
• Founding-team stability. Whether the Ho, Balsam, and McGrath team remains intact through the post-Series B scaling phase, given the 2025 to 2026 pattern of senior researchers leaving frontier laboratories to found independent ventures.
• Frontier-lab competitive response. Whether Anthropic, OpenAI, or Google DeepMind launch competing standalone interpretability products, or whether the in-house interpretability programs at frontier laboratories continue to treat Goodfire as a complementary commercial provider.
• Second-act trajectory. Whether Goodfire produces a more significant scientific or commercial outcome than the original RippleMatch venture, and whether the underlying interpretability thesis holds at the larger valuations the Series B implies.

Sources

• Eric Ho LinkedIn profile. Public profile listing the Goodfire and RippleMatch roles, the Yale University education, and the Forbes 30 Under 30 recognition.
• Eric Ho on LinkedIn: I made the difficult decision to leave RippleMatch. The December 2023 LinkedIn announcement of the RippleMatch transition and the AI-safety pivot.
• Some for-profit AI alignment org ideas. The December 2023 LessWrong essay outlining the case for commercially-funded AI safety organizations.
• Yale startup matches students to jobs. The February 2019 Yale Daily News profile of RippleMatch listing Eric Ho '16 as co-founder.
• RippleMatch Cofounders Recognized on Forbes 30 Under 30 List in Enterprise Technology. The 2022 Forbes 30 Under 30 announcement.
• RippleMatch Raises $45m from Goldman Sachs Asset Management. The June 2022 Series B announcement.
• Goodfire Raises $7M to Break Open the Black Box of Generative AI Models. The August 2024 seed-round announcement.
• Announcing Our $50M Series A to Advance AI Interpretability Research. The April 2025 Series A blog announcement, with Ho's quote on the mechanisms by which AI models fail.
• Anthropic backs Goodfire in $50M Series A funding to decode AI models, marking first-ever startup investment. The Tech Startups coverage of Anthropic's first-ever startup equity investment.
• Leading Goodfire's $50M Series A to Interpret How AI Models Think. The Menlo Ventures partner perspective on the Series A, including Deedy Das's lead-investor role.
• Understanding, Learning From, and Designing AI: Our Series B. The February 2026 Series B blog announcement, with Ho's quote on interpretability as a toolset for a new domain of science.
• AI Lab Goodfire Raises $150M at $1.25B Valuation to Design Models with Interpretability. The Series B press release.
• Startup Goodfire Notches $1.25 Billion Valuation to Decode AI Models. Bloomberg coverage of the Series B valuation.
• Goodfire's Eric Ho on Mapping the Mind of a Neural Net. The July 2025 Sequoia Capital Training Data podcast episode hosted by Sonya Huang and Roelof Botha.
• Mapping the Mind of a Neural Net: Goodfire's Eric Ho on the Future of Interpretability. YouTube recording of the Sequoia Training Data episode.
• Goodfire Business Breakdown & Founding Story. Contrary Research dossier covering the founding story, the December 2023 LessWrong essay, and the Daniel Balsam and Tom McGrath co-founder backgrounds.
• Goodfire Ember on GitHub. The public software development kit for the Ember platform.