Edinburgh AI fintech Aveni banks £12m to police the conduct of finance's new robot workforce
Lloyds- and Nationwide-backed Aveni has closed a £12 million round led by PXN Ventures, doubling down on tools designed to keep AI agents inside the lines drawn by UK regulators.
Edinburgh-based artificial intelligence specialist Aveni has closed a £12 million funding round, in a deal that hands the Scottish capital one of the most credible homegrown answers to the question now haunting the City: who is watching the machines?
The round was led by PXN Ventures, the fast-growing investment house operating outside the gravitational pull of London and the South East. Existing backers Puma Growth Partners, Lloyds Banking Group, Nationwide and Scottish Enterprise all returned to the cap table, an unusually clean signal of confidence in a market where follow-on cheques have grown harder to win.
The proceeds will accelerate Aveni's Unified Assurance Platform and the launch of two new products, Agent Assure and Agent Approve, both engineered to assess the conduct risk of AI agents that talk directly to consumers in regulated financial services.
Banks, insurers and wealth managers are racing to deploy generative AI inside customer journeys that the Financial Conduct Authority polices line by line. Aveni's own research suggests just two per cent of firms believe they have adequate guardrails in place. The gap between ambition and assurance is, in effect, the company's market.
Founded seven years ago and now a fixture of Scotland's growing technology cluster, Aveni already counts banks and large wealth managers among the users of Aveni Detect, its AI compliance monitoring tool, and Aveni Assist, a productivity layer for advisers. Both products sit on top of FinLLM, the firm's proprietary suite of small language models trained on United Kingdom financial services data — a deliberately British answer to the question of whose data is shaping the models that read British conversations.
The new Agent Assure product is the centrepiece of the raise. It extends the same conduct-monitoring logic Aveni has long applied to human advisers into the world of autonomous agents, allowing firms to track how an AI behaves in a consumer interaction in the same dashboard that monitors a human call handler.
The FCA has been blunt that consumer outcomes must be assessed consistently, whether the interaction is run by a person or a piece of software. Aveni is a participant in the regulator's Supercharged Sandbox programme and has set up an Agent Assurance Expert Council to help shape what good practice looks like before policy is set in stone.
Chief executive Joseph Twigg said the continued backing of existing investors was a powerful endorsement of the firm's direction. "We have spent seven years building the models, the experience and the regulatory relationships that make us uniquely qualified to solve the hardest problem in AI adoption right now: how do you assure the conduct of an AI agent interacting with a real customer," he said.
Industry watchers expect a wave of consolidation in the AI assurance niche over the next eighteen months as larger software houses look to bolt conduct-monitoring onto their offerings. Aveni's combination of FCA relationships, deployed products inside two of the UK's biggest high-street lenders and a Scottish base far from the London tech salary spiral may make it an awkward target to copy and an attractive one to partner with.
For the wider United Kingdom economy, the round is another data point in a year that has seen domestic technology firms attract record capital despite a tougher investor filter. Scotland in particular continues to punch above its weight; Aveni's £12 million round will not crack any unicorn league tables, but in a sector where credibility is built deployment by deployment, it is the type of cheque that quietly compounds.