Y Combinator is supported by the base of the rules

Y combinator-alum Rules It is betting that the next wave of automation in financial services will not revolve around the cheerful artificial intelligence facades, but the tasks of the unpracted back office like compliance.

The startup company, founded by Ebose and Chidi Williams, the Nigerian engineers who met in London, collected a 2.1 million dollar researcher, led by Bowry Capital, with the participation of Y Combinator, Commerce Ventures, the Transpose VC platform, along with many angels.

Financial services companies spend huge amounts of effort on support tickets, conflict resolution, guarantee guarantee, and organizational compliance. Rolebase, which is called an agent colleague, is replaced by many handicrafts in these tasks. The artificial intelligence agent can assess customer interactions, organizational risks to the media, and leads to correct follow -up through tools such as Zendesk, Jira and Slack without losing human supervision in the episode required by financial companies.

“Our co -worker” integrates through platforms and cooperates with human agents and the back office teams to completely manage the dispute’s dispute life cycle with time saving, reduce errors, and maintain compliance. ” Currently, the start of a year ago has been published on clients such as the American Banking Banking platform and a financial institution from Fortune 50 whose name has not been revealed.

The base of the rules was not the first swing of the founders. EBose, a previous company for Microsoft and Williams, a previous engineer at Goldman Sachs, built many products together like an artificial intelligence customer notes tool before eventually settled on the base of the rules. The idea came after seeing how the operations of the background office were ineffective in small and large financial institutions, especially when it comes to the progress of organizational work.

The emerging company is currently focusing on the workflow resulting from customer service reactions, with the first wedge on quality guarantee. Quality assurance analysts in traditional financial institutions usually review 3-5 % of support reactions to ensure follow-up of compliance protocols for actors.

The founders say that the rule of rules now evaluates 100 % of these reactions, which reduces costs by up to 70 %. In the case of RHO, for example, the Rulebase base helped reduce escalation by up to 30 %.

TECHRUNCH event

San Francisco
|
27-29 October, 2025

“We are performing the automation of the workflow that begins with customer interaction, and the areas we are doing already in dealing with a comprehensive to a party,” Ebose CEO said in an interview with Techcrunch. “Although many of this is the guarantee of quality, compliance and conflicts associated with calls and customer messages, our long -term goal is to take the largest possible number of manual rear office tasks by withdrawing these steps and segmented tabs to one coordinated workflow.”

The new financing will help him to weaken engineering and ultimately add new features to a classmate of artificial intelligence such as fraud, preparation, and organizational reporting.

Rolebase’s base focuses on financial services today because automation requires accuracy. Ebose said: “You need to understand MasterCard rules, CFPB schedules. This is the depth of the domain knowledge of our ditch.”

The company targets commercial banks, new exporters, and cards across Africa, Europe, and the United States, but the road map can ultimately include neighboring sectors such as insurance, where there is a similar workflow.

The founders say the revenues are growing rapidly, with “two numbers” growth for a month since it joins the Yombinator’s Fall 2024 batch. The Rulebase Base Model is used, or charging each reaction that is reviewed or automated workflow.

As one of the few African founders who enter into building AI tools, Ebose and Williams advice for the founders who are trying to accept it in the global accelerator is thinking globally since the first day.

“We are in a moment when the small teams can provide more value, more quickly, more than ever, so the meant yourself on” X for Y “or vertical narrow looks like a lost chance.” “The basis for the tendency to the text with more than 300,000 downloads,” Williams, who before the base, said, “Williams, who built the tanner, an early tool from speaking to the text with more than 300,000 downloads.