Founder · Enterprise AI · New York
Conor Twomey
He left a Managing Director title to solve the quiet problem behind most failed AI projects: the machines are smart, but they don't understand the business. AI One is his answer.
He left a Managing Director title to solve the quiet problem behind most failed AI projects: the machines are smart, but they don't understand the business. AI One is his answer.
Conor Twomey has spent his career close to data that mattered - the kind that moved money at Citi, Goldman Sachs and Morgan Stanley, and the kind that shaved fractions of a second off a Formula One pit stop. Somewhere across those fifteen years, he noticed a pattern that would eventually become a company. Enterprises kept pouring money into data infrastructure, and their artificial intelligence kept coming back with confident, expensive guesses.
The problem, as he came to see it, was not raw intelligence. The models were capable. What they lacked was context - a grounded understanding of the specific business they were meant to serve. A model can write fluent prose about a sales pipeline without knowing which deals are real, which systems hold the truth, or who is allowed to see what. That gap, Twomey argues, is where most corporate AI quietly falls apart.
In 2024 he acted on the observation. Together with fellow Irish entrepreneur Fergus Keenan, he co-founded AI One in New York and took the chief executive role. The company describes itself as the infrastructure layer that lets enterprise AI understand business context. In plainer terms, it connects directly to the systems a company already runs - Salesforce, Workday, on-premise databases - and resolves their fragmented data into something an AI model can reason over, without first demanding a costly migration to a shiny new platform.
AI can only perform as well as the context it understands. We created AI One to give enterprises control over that context so their AI isn't guessing, it's operating with understanding.
His roots are academic and quantitative. Twomey earned a BSc in Financial, Mathematical and Actuarial Sciences from University College Cork, then an MSc in Computational Finance from the University of Limerick. It is an unusually numerical foundation for a founder now selling to boardrooms, and it shows in how he frames the pitch - less about revolution, more about measurable arithmetic.
Before AI One, he spent roughly seven years at the data-analytics company KX, climbing through professional-services and presales roles to become Head of AI Strategy and Managing Director for North America. In that seat he helped steer KX toward being a recognised enterprise AI platform and struck partnerships with Microsoft, AWS, Google Cloud, NVIDIA, Intel, Databricks and Snowflake. It was, in effect, a front-row seat to every reason enterprise AI stalls.
Twomey does not soften his diagnosis. He estimates that around 85% of enterprise AI projects fail, and he traces the failures to a single, unglamorous cause: nobody can point to the line item where cost dropped or revenue rose. Companies, he says, are "building monuments, not machines" - erecting impressive data architecture while the outcome, the actual working system, never ships.
His prescription is deliberately narrow. Rather than fund another data-modernisation slide deck, he tells prospective clients to pick a single workflow where seconds matter, and start there. AI One leans into speed as its proof: a working prototype in about a week, a limited production system in roughly five weeks, and a full production deployment in about ten. The promise is not that AI will transform everything, but that it can do one important thing well, quickly, and provably.
For all the automation talk, Twomey is firm about keeping people in the loop. His guiding rule is that every automated decision must have a human who can explain it, reverse it, and be accountable for it. He rejects the replacement narrative, framing AI instead as something that "absorbs the repetitive keyboard work" so that talented people can move to strategy and judgment.
He also champions what he calls "AI-shoring" - bringing critical workflows back in-house while letting autonomous systems handle the drudgery, blending a security argument with an efficiency one. It is a founder's worldview shaped by years inside regulated, high-stakes industries where a wrong answer is not a novelty but a liability.
Before you fund another data-modernisation slide deck, pick a workflow where seconds matter.
The market has responded with capital. AI One raised $11 million, including a $7 million Series A led by Vestigo Ventures, whose general partner Mark Casady called the platform's ability to activate enterprise context without migration "a breakthrough." Beyond the company, Twomey keeps a public profile - a 2024 Wall Street 50 honoree, a World Economic Forum Global Shaper, and a US board member of the humanitarian organisation GOAL Global.
What comes through, across interviews and podcasts, is a founder allergic to hype selling a product built on hype's shortcomings. He talks about precision, accountability and outcomes in a field addicted to scale and spectacle. Whether AI One becomes the standard layer for enterprise context is unwritten. The thesis, at least, is clear: give the machine an understanding of the business, and it stops guessing.
Source: SiliconRepublic, Businesswire, company disclosures
AI isn't an arms race; it's a precision strike.
Turn AI from a slide-deck promise into operational reality.
Every automated decision must have a human who can explain it, reverse it, and be accountable for it.
The budget owner can't point to a line item where cost dropped or revenue rose.
Turn employees into AI power users, not spectators.
Most companies are building monuments, not machines.
Built AI One into a funded enterprise AI company in under two years.
Including a $7M Series A led by Vestigo Ventures.
2024 honoree, Irish America Magazine.
Plus a US board seat at humanitarian organisation GOAL Global.
VV FinTech Podcast · Delivering AI Solutions to the Enterprise
▶ Watch on YouTube