Record OS

Dhruv Chadha studied at the University of Oxford before building a career as a finance executive at some of the most analytically demanding institutions in technology. He worked in finance roles at Google before joining Wise, where he operated inside one of the most data-driven financial technology organisations in the world and witnessed firsthand how payment complexity creates cascading complexity in personal tax situations for employees with equity, multi-currency income, and international earnings. He then became CFO at Nansen, the blockchain analytics platform, which placed him at the intersection of crypto finance and the specific tax treatment problems that arise when income is denominated in volatile digital assets. That is three consecutive roles where the people around him had income structures that made self-assessment genuinely difficult. While testing the Record OS platform, Dhruv found a £38,968 error on his own 2023-2024 Self Assessment, even with professional help. He posted about this publicly on LinkedIn. The willingness to make that admission public is a specific kind of intellectual honesty: he is not selling a product that claims to be perfect. He is selling a product that caught an error that a human professional missed, in his own return, and whose output is better because the AI and the accountant check each other. That is a more credible product thesis than any benchmark. His ZoomInfo profile carries a reference quote from a colleague: "Dhruv is very detail-oriented and while working together, he watched the balance sheet like a hawk without ever losing sight of our primary objective." That is the CFO character. The balance sheet precision of a professional accountant combined with the systems thinking of a technology operator is exactly the combination needed to design a product that is both technically sound and financially credible with accounting professionals. He met José in mid-2024 and the two spent time experimenting with AI applications before deciding that self-assessment was the category worth building into. They began building Record OS at the start of 2025. His public LinkedIn output since founding concentrates on the MTD IT policy timeline, the self-assessment error rate problem, and product updates from the beta. He has also posted about Broc, which appears to be a prior product experiment from his Wise period. He operates in the UK fintech community and attended the Accountex conference to represent Record OS, which is the correct venue for reaching accounting firm decision-makers.

José Luis De La Peña holds a Master of Science in Telecommunications Engineering from Danmarks Tekniske Universitet (DTU) in Denmark (2010-2012), one of the leading technical universities in Scandinavia, which gave him his foundation in distributed systems, signal processing, and network architecture before he moved into software engineering professionally. His career moved from early freelance and agency engineering work through SubscribEz and Reto-Moto before he joined Generation Home (GNH), a London-based mortgage lender that was building digital mortgage infrastructure for UK homebuyers, as an engineer. GNH is the mortgage lender that Record OS's brief refers to: the experience of building financial services software for a compliance-heavy, document-intensive industry gave him exactly the engineering context that self-assessment automation requires. He then joined Wise (then TransferWise) as a software engineer, where he worked on systems at a company that had already redefined how international payments worked and was preparing for its Nasdaq listing. His LinkedIn post on the day of the Wise Nasdaq listing reflects genuine pride in what the team had built. The experience of building engineering infrastructure at a regulated fintech that operates under multiple national regulatory frameworks is the preparation for building an HMRC-integrated AI platform. His early research work, visible in his LinkedIn profile, involved integrating social network APIs into a unified graph using NoSQL technologies like CouchDB and Redis, which is an early signal of his interest in data integration and unification problems — exactly the engineering challenge at the core of Record OS, where financial data from multiple sources (bank feeds, accountant records, HMRC data) must be normalised into a single coherent filing. He moved from computer science and legal tech into engineering roles at Wise and mortgage lender GNH. The legal tech reference in his background, not confirmed in publicly available sources but noted in AccountingWEB, suggests he has exposure to document-heavy, compliance-sensitive software environments beyond just fintech. His DTU telecom background, his GNH fintech engineering, and his Wise payments experience together make him the right technical co-founder for a product that needs to integrate cleanly with HMRC APIs, handle complex financial document parsing, and operate under the audit and data security expectations of UK accounting professionals.

Dhruv Chadha studied at the University of Oxford before building a career as a finance executive at some of the most analytically demanding institutions in technology. He worked in finance roles at Google before joining Wise, where he operated inside one of the most data-driven financial technology organisations in the world and witnessed firsthand how payment complexity creates cascading complexity in personal tax situations for employees with equity, multi-currency income, and international earnings. He then became CFO at Nansen, the blockchain analytics platform, which placed him at the intersection of crypto finance and the specific tax treatment problems that arise when income is denominated in volatile digital assets. That is three consecutive roles where the people around him had income structures that made self-assessment genuinely difficult. While testing the Record OS platform, Dhruv found a £38,968 error on his own 2023-2024 Self Assessment, even with professional help. He posted about this publicly on LinkedIn. The willingness to make that admission public is a specific kind of intellectual honesty: he is not selling a product that claims to be perfect. He is selling a product that caught an error that a human professional missed, in his own return, and whose output is better because the AI and the accountant check each other. That is a more credible product thesis than any benchmark. His ZoomInfo profile carries a reference quote from a colleague: "Dhruv is very detail-oriented and while working together, he watched the balance sheet like a hawk without ever losing sight of our primary objective." That is the CFO character. The balance sheet precision of a professional accountant combined with the systems thinking of a technology operator is exactly the combination needed to design a product that is both technically sound and financially credible with accounting professionals. He met José in mid-2024 and the two spent time experimenting with AI applications before deciding that self-assessment was the category worth building into. They began building Record OS at the start of 2025. His public LinkedIn output since founding concentrates on the MTD IT policy timeline, the self-assessment error rate problem, and product updates from the beta. He has also posted about Broc, which appears to be a prior product experiment from his Wise period. He operates in the UK fintech community and attended the Accountex conference to represent Record OS, which is the correct venue for reaching accounting firm decision-makers.

José Luis De La Peña holds a Master of Science in Telecommunications Engineering from Danmarks Tekniske Universitet (DTU) in Denmark (2010-2012), one of the leading technical universities in Scandinavia, which gave him his foundation in distributed systems, signal processing, and network architecture before he moved into software engineering professionally. His career moved from early freelance and agency engineering work through SubscribEz and Reto-Moto before he joined Generation Home (GNH), a London-based mortgage lender that was building digital mortgage infrastructure for UK homebuyers, as an engineer. GNH is the mortgage lender that Record OS's brief refers to: the experience of building financial services software for a compliance-heavy, document-intensive industry gave him exactly the engineering context that self-assessment automation requires. He then joined Wise (then TransferWise) as a software engineer, where he worked on systems at a company that had already redefined how international payments worked and was preparing for its Nasdaq listing. His LinkedIn post on the day of the Wise Nasdaq listing reflects genuine pride in what the team had built. The experience of building engineering infrastructure at a regulated fintech that operates under multiple national regulatory frameworks is the preparation for building an HMRC-integrated AI platform. His early research work, visible in his LinkedIn profile, involved integrating social network APIs into a unified graph using NoSQL technologies like CouchDB and Redis, which is an early signal of his interest in data integration and unification problems — exactly the engineering challenge at the core of Record OS, where financial data from multiple sources (bank feeds, accountant records, HMRC data) must be normalised into a single coherent filing. He moved from computer science and legal tech into engineering roles at Wise and mortgage lender GNH. The legal tech reference in his background, not confirmed in publicly available sources but noted in AccountingWEB, suggests he has exposure to document-heavy, compliance-sensitive software environments beyond just fintech. His DTU telecom background, his GNH fintech engineering, and his Wise payments experience together make him the right technical co-founder for a product that needs to integrate cleanly with HMRC APIs, handle complex financial document parsing, and operate under the audit and data security expectations of UK accounting professionals.