Eiso Kant
Last reviewed
Jun 8, 2026
Sources
14 citations
Review status
Source-backed
Revision
v1 · 1,611 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 8, 2026
Sources
14 citations
Review status
Source-backed
Revision
v1 · 1,611 words
Add missing citations, update stale details, or suggest a clearer explanation.
Eiso Kant is a Dutch software entrepreneur and technology executive who is a co-founder and the chief technology officer of Poolside, a company building foundation models for software development. He co-founded Poolside in 2023 with Jason Warner, the former chief technology officer of GitHub, and the two are now generally described as co-founders and co-chief executives, with Kant continuing to lead the company's technical research.[1][2] Before Poolside, Kant spent more than a decade founding developer-focused companies, most notably source{d}, an early entrant in applying machine learning to source code, and Athenian, an engineering-analytics platform.[2][3]
Kant is Dutch and was raised in the Netherlands, where he attended international schools from around the age of ten and began programming as a child.[4] He started building businesses while still a teenager. At about fifteen he launched PrimaPrints, an e-commerce site selling classic eighteenth and nineteenth century art prints, a project he has described as a months-long exercise in teaching himself web design and programming.[4]
In early 2009 Kant co-founded Twollars with the telecommunications analyst Mac Taylor. Twollars billed itself as the first virtual currency on Twitter and was used mainly to direct donations to charities; according to Kant it signed up more than 300 charities across over 30 countries, including the Red Cross and Livestrong, before he wound it down to continue his studies.[4] He went on to earn a bachelor's degree in business administration from IE University (Instituto de Empresa) in Madrid, Spain.[2][4]
While at IE University, Kant co-founded the online-recruitment startup Tyba in 2011 together with Jorge Schnura and Philip von Have. Tyba built personal profile pages that let junior technology talent in Europe present themselves to employers beyond a traditional resume. The company was acquired by the Danish recruitment platform Graduateland in April 2016.[2][3]
| Venture | Years | Kant's role | Outcome |
|---|---|---|---|
| PrimaPrints | mid-2000s | Founder | Early personal project |
| Twollars | 2009 | Co-founder | Wound down |
| Tyba | 2011 to 2016 | Co-founder | Acquired by Graduateland (2016) |
| source{d} | 2015 to 2019 | Co-founder and CEO | Wound down |
| Athenian | 2019 to 2023 | Founder and CEO | Acquired by the Linux Foundation (2023) |
| Poolside | 2023 to present | Co-founder and CTO | Active |
In 2015 Kant co-founded source{d} in Madrid with the same partners from Tyba, Jorge Schnura and Philip von Have. The company set out to apply machine learning to large bodies of source code, a field Kant has described as among the first dedicated efforts to apply artificial intelligence to code and software.[2][5] source{d} built open infrastructure for retrieving, parsing, and analyzing code at scale, drawing on neural-network techniques applied across millions of public repositories. Press coverage from 2018 reported that its analysis spanned on the order of 17 million repositories and several million developers.[5]
The company released tooling under permissive open-source licenses, including the source{d} Engine for large-scale code analysis and source{d} Lookout for assisted code review, and it became known for developing in the open, publishing not only its code but elements of its internal operations.[5] source{d} raised roughly 10 million dollars in venture funding, including an approximately 6 million dollar Series A in October 2016 backed by investors such as Xavier Niel, Otium Venture, and Sunstone Capital.[2][6] After shipping an enterprise edition in 2019, the company wound down its operations.[3] Kant served as chief executive officer throughout source{d}'s life.
In December 2019 Kant founded Athenian, an engineering-analytics platform that gave software teams metrics on their development process, such as cycle time and pull-request flow. A defining choice was Athenian's focus on team-level and process-level measurement rather than ranking individual engineers, a deliberate departure from the per-developer scoring common in the category.[7] Kant led the company as chief executive until 2023, when he left to start Poolside; the Athenian platform was subsequently acquired by the Linux Foundation later that year.[2][3]
Kant and Jason Warner co-founded Poolside in 2023. The two had co-hosted a podcast on engineering leadership, Developing Leadership, since 2020, and the idea for a frontier AI company focused entirely on code, software, and version-control history grew out of those conversations.[1][8] At founding, Warner took the role of chief executive officer and Kant the role of chief technology officer; by 2025 and 2026 both founders were widely described as co-chief executives, with Kant continuing to lead Poolside's technical and research work.[1][2]
Poolside is incorporated in the United States and headquartered in San Francisco, with a substantial team in Paris and elsewhere. The Paris presence led some coverage to describe it as a French company, a characterization Kant has publicly rejected, stating that Poolside is an American company with a globally distributed team.[9] By late 2025 the company employed roughly 150 people.[10]
The company raised money quickly. An initial 26 million dollar seed round led by Redpoint Ventures closed in May 2023, and an extension in August 2023 brought the seed to about 126 million dollars at a valuation near 526 million dollars, led by Xavier Niel and Felicis with participation from Bpifrance, Motier Ventures, and others.[8][9] In October 2024 Poolside closed a 500 million dollar Series B led by Bain Capital Ventures at a valuation of about 3 billion dollars, with participation from Nvidia, eBay Ventures, DST Global, StepStone Group, Citi Ventures, Felicis, and Redpoint, bringing total funding to roughly 626 million dollars.[1][11][12]
In October 2025 Nvidia announced plans to invest up to 1 billion dollars in Poolside as part of a roughly 2 billion dollar round that would have valued the company at about 12 billion dollars, and Poolside partnered with CoreWeave on Project Horizon, a planned large data-center campus in West Texas.[12] Those plans came apart in early 2026: the round did not close, and in April 2026 CoreWeave terminated its anchor lease for the campus, leaving the company's 2024 Series B as its most recent completed raise.[13] As of mid-2026 Kant remains a co-founder and co-chief executive of an independent Poolside; he is based in Europe.[2]
Under Kant's technical leadership, Poolside concentrates its data, training, and infrastructure entirely on software development rather than treating code as one capability among many. Kant and Warner have framed the goal as reaching a form of narrow artificial general intelligence through software and code, arguing that a model specialized for programming can surpass general-purpose large language models on engineering tasks.[1][2]
The central technical idea Kant has championed is Reinforcement Learning from Code Execution Feedback, abbreviated RLCEF. Rather than learning only by reading existing source code, Poolside's models attempt realistic coding tasks against real code bases, then compile and run the resulting programs and execute tests, using the outcomes as a reward signal in a large-scale reinforcement learning loop. The aim is to mimic how human engineers actually improve, by writing programs and learning from whether they run correctly.[1][14] Generating that feedback at scale is infrastructure-heavy: Poolside has described running on the order of a million container images and thousands of code executions per minute to produce synthetic training data.[14]
Poolside packages this work in an internal pipeline it calls the Model Factory and has released models including Malibu, aimed at complex multi-step engineering tasks, and Point, a smaller model tuned for fast code completion. Consistent with Kant's enterprise focus, the models are offered for deployment inside a customer's own infrastructure so that proprietary code does not leave the organization, an approach intended to appeal to large enterprises and public-sector users with strict data-control requirements.[14] Kant is also a frequent public speaker on AI code generation, and he has argued that most companies should not attempt to build their own foundation models.[2][10]