AI agents have fast become one of the most hotly-debated solutions in tech. According to PWC, 88% of senior executives plan to increase AI budgets due to agentic AI, with 79% saying that AI agents are already being adopted in their companies.
London-based startup Gradient Labs, founded by the team that previously led AI and data initiatives at Monzo, has fast become one of the most promising agentic AI startups in the UK. Its solution provides banking customers with access to AI agents, which can act as dedicated account managers.
In July 2025, the company announced a $13 million USD (€11.10 million) Series A funding round led by Redpoint ventures, after building what it claims is the first compliant AI platform for the financial services industry. After this funding round, Sifted has since placed the company’s valuation at $60 million USD (€51.1 million).
The company developed traction in a financial services industry that is often resistant to change, with customers including Wise, SteadyPay, Digital Bank, Zego, and Plum. Part of its success has come down to focusing on streamlining customer support.
Scaling like only an AI-first company can
According to Gradient Lab’s team, the two-year-old startup scaled with a team of less than 100 people and delivered 10x revenue growth over the past year – success it attributes to its use of agentic AI to automate complex customer service queries and deliver higher customer satisfaction scores than most human teams.
Gradient Labs can be considered an AI-first startup, using AI agents to automate core business processes. “We have no specialist teams so far. For example, we don’t have full-time HR, finance, or legal teams. While we outsource some tasks, other activities are already picked up by AI agents,” said Dimitri Masin, co-founder and CEO at the startup.
“This structure, flatter, more agile, is how most big companies will operate in the near future. The trend is industry-wide, because it allows companies to scale with more purpose.”
Masin claims that using internal support tools to onboard new customers and leveraging AI that learns from human expertise helped to scale the company’s revenue tenfold over the last year while keeping headcount flat. He notes that the startup isn’t necessarily aiming to be a 200-person company, as leveraging talent isn’t just about the number of employees anymore.
“Thanks to AI agents, we can focus our hiring budget on headhunting absolute AI-players, not average paper-pushers. In the upcoming ‘talent war,’ one person with the right expertise plus an AI agent can leverage more impact than five people with some basic understanding of the field,” Masin added.
In this sense, AI-first companies have access to scalability that more traditional organizations do not; agents can be used to automate workflows to augment the capabilities of human staff in a way that could be a significant force multiplier.
It also presents an alternative to the inflexible attitude of larger organizations, locked into outdated SaaS solutions and fixed workflows which stifle agility. Modern startups need to move fast if they want to be able to compete.
Using AI agents to streamline customer support
In highly-regulated sectors like finance, AI agents introduce new risks that banks and financial service institutions need to be prepared to address. While banks can develop their own AI agents and deploy them at considerable cost, Gradient Labs provides access to a prebuilt no-code AI agent with out-of-the-box guardrails.
One of its agents, the Borrower Lifecycle Agent, is designed specifically to support lending, covering the borrower’s journey from applications to collections. Once the provider approves the loan, the agent can make a personalised welcome call, walk the customer through a payment schedule, set up autopay,and provide details on customer support options.
When payments are overdue, the agent can reach out to the end customer and explain what they owe. In this context, the agent knows the borrower’s balance, payment history and repayment options in each of its interactions. The appeal here is scalability.
“Whenever you run a large consumer organization, a significant part of what makes it work is customer operations,” Masin said. “We enable banks to scale far more efficiently. For example, where they might have needed 5,000 people before to run manual workflows like the account and transaction reviews, KYC, AML screening, etc., they can now operate with two to five fewer staff. So they can use their own staff to achieve much higher goals.”
Another key selling point of Gradient Lab’s approach is its built-in compliance. Each agent runs over 20 financial services-specific guardrails. Agents interacting with UK and EU customers, for instance, are FCA Consumer Duty, CONC, vulnerable customer handling, Breathing Space, GDPR, and EU AI Act compliant.
Making agent pilots from scratch is extremely challenging, as evidenced by the fact that 95% of generative AI pilots at companies are failing to deliver noticeable returns. Yet, providers like Gradient Labs that sell practical implementation are thriving – precisely because their DIY approaches are difficult to implement.
