EN
Back to the archive

The archive · Money & Fintech · Product decision · 2023-2026

Gradient Labs' bet: Monzo alumni run bank ops with agents; Series A hits $26M

Founded by Monzo's AI team, Gradient Labs sells compliance-native agents for lending, disputes and KYC; revenue rose 900% and its Series A doubled to $26M.

Gradient Labs

The betDomain-specific agents with compliance built in can autonomously run lending, disputes and KYC, replacing manual operations inside regulated finance.Scaling

What the business is

Gradient Labs builds vertical AI agents for financial services that automate customer servicing, lending lifecycles, dispute resolution, KYC checks and collections inside a bank's existing systems, with guardrails mapped to rules such as FCA Consumer Duty and the EU AI Act.

Starting capital$13M initial Series A (July 2025, led by Redpoint Ventures), upsized by $13M in June 2026 with Octopus Ventures and CommerzVentures joining, bringing total funding to $42.6M

How it started

Gradient Labs was founded in 2023 in London by Dimitri Masin, Neal Lathia and Danai Antoniou, early employees of challenger bank Monzo who helped build its data science and machine learning capabilities across customer operations and financial crime. Having seen how hard regulated customer operations were to manage at scale, they concluded that compliance-constrained finance needed purpose-built agents rather than generic AI tools.

What happened

The company launched agents for lending, disputes and KYC, integrating them into banks' existing infrastructure so institutions could automate workflows without ripping out their systems. It raised an initial $13M Series A in July 2025 led by Redpoint Ventures, with LocalGlobe, Puzzle Ventures, Liquid 2 Ventures and Exceptional Capital participating. On May 31, 2026 it announced a $13M extension led by Octopus Ventures and CommerzVentures, doubling the Series A to $26M, as revenue grew 900% over the previous year and customers expanded on both sides of the Atlantic.

How it ended up

With the upsized round, Gradient Labs was building toward autonomous banking, automating complex, long-running financial operations from customer servicing to collections, and preparing to expand across the United States with its Lending, Disputes and KYB agents.

Background

Gradient Labs is a London startup building vertical AI agents for financial services. Founded in 2023 by three early Monzo employees, it automates the most expensive and time-consuming operations in regulated finance, including customer servicing, lending workflows, dispute resolution, KYC checks and collections, working inside the systems banks already run.

The founding team, Dimitri Masin, Neal Lathia and Danai Antoniou, had built Monzo's data science and machine learning capabilities across customer operations and financial crime. Their thesis was that compliance-constrained finance would not trust generic AI: each Gradient Labs agent, such as the Lending Agent, Disputes Agent or KYB Agent, carries guardrails and test scenarios mapped to rules including FCA Consumer Duty and the EU AI Act.

Gradient Labs raised an initial $13M Series A in July 2025 led by Redpoint Ventures, then announced on May 31, 2026 a $13M extension led by Octopus Ventures and CommerzVentures, doubling the round to $26M. At that point revenue had grown 900% over the previous year, customer satisfaction scores reached up to 98%, and its lending agents alone handled hundreds of thousands of customer calls monthly.

The company counted Wise, Monzo, Zego and Pockit in Europe, plus Current, Stash and Rho in the United States, collectively reaching more than 32 million end users. The bet is that banks will move from bolt-on AI assistants to agents that autonomously execute operational tasks inside financial systems, and that domain specialization, not model size, is what makes that safe enough for regulators.

What has to be true

  • The founders built the exact systems they now automate while at Monzo, giving them credibility with banks and an unusually precise map of regulated customer operations.
  • Each agent ships with guardrails and test scenarios mapped to rules like FCA Consumer Duty and the EU AI Act, making regulation part of the product rather than a sales obstacle.
  • The platform integrates with existing bank infrastructure instead of demanding rip-and-replace, so institutions can automate workflows without a multi-year migration.
  • Named customers across both the UK and US, 32 million end users and 900% revenue growth gave investors hard evidence before the Series A was doubled.

What can be applied

In regulated industries, encode compliance into the agent itself: domain guardrails and test scenarios become the moat that generic copilots cannot cross.

Aftermath

After the June 2026 extension, Gradient Labs planned to expand across the United States and keep building autonomous banking tools that handle long-running processes from missed-payment collections to agreed repayment plans. Its positioning shifted from conversational AI to agents that execute operational tasks directly inside financial systems, competing with general customer-service platforms by staying narrowly focused on regulated finance.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases