Kastle Raises $24 Million to Deploy AI Agents Across Legacy Banking Cores

Kastle has raised a $24 million Series A round led by Insight Partners to build and scale autonomous AI agents for consumer lending and mortgage servicing. The round includes continuing backing from Y Combinator and Commerce Ventures, along with new investment from Fifth Wall and industry executives, as reported by Pulse 2.0 and Beinsure.

The capital comes as the company expands from early customer workflows into wider enterprise lending operations. Kastle states that its AI agents have processed more than $1.8 billion in transactions for major banks and lending institutions.

What Kastle Builds

Kastle builds an AI workforce layer tailored specifically to financial institutions, targeting high-friction operations like mortgage servicing and loan origination. In financial services, core infrastructure often runs on decades-old software that is difficult and costly to replace. Rather than asking lenders to execute multiyear core platform migrations, Kastle deploys agents directly on top of existing environments.

The technical stack is built on Python, large language model backends, and transformer-based voice models. These systems interface directly with legacy core banking databases, automate manual workflows, handle customer calls, screen leads, and keep institutional systems of record synchronized.

In an interview with PYMNTS.com, Chief Executive Officer Rishi Choudhary described the product as "the action layer on top of" legacy cores. Under this setup, software agents take on specific tasks—such as inbound servicing inquiries, payment collections, or case files—within rigid administrative rules. If a borrower requests an action outside an agent’s permitted authority, such as a payment reversal, the system routes the request to human managers or authorized escalation staff.

Leadership Behind the System

Kastle's executive group pairs software product and infrastructure experience with operational leadership.

Co-founder and CEO Rishi Choudhary studied systems engineering at the University of Illinois Urbana-Champaign (UIUC), where he ran a 250-student consulting organization with $500,000 in annual recurring revenue. Before founding Kastle, he served as founding Product Manager for Redfin's Mortgage Marketplace at age 22, doubling that business from $3 million to $6 million in revenue across nine months, and led Redfin's Search Experience team responsible for 80 percent of the site's traffic.

Co-founder and Chief Technology Officer Nitish Poddar also graduated with a degree in computer science from UIUC. Poddar was previously a tech lead at Verkada, where he developed a patented motion search system, designed the company’s backend inference infrastructure, ran the firmware delivery team for a fleet management system, and led engineering for Verkada's Incident Management line.

The company's corporate leadership records also list Chief Executive Officer Haniel Lynn, a University of Pennsylvania engineering graduate, former McKinsey & Company consultant, and former Group President at the Corporate Executive Board, alongside Co-chairman Piyush Sodha, former CEO of Cibernet Corporation.

The Strategy: Automating Workflows Without Ripping Out Cores

Large financial institutions face severe bottlenecks when rolling out automated software. As Insight Partners Managing Director Rebecca Liu-Doyle stated to Pulse 2.0, lenders do not need software layers that add administrative overhead; they need systems capable of handling strict compliance and completing high-stakes work without multiyear IT overhauls.

As Finextra Research reported, Kastle operates on a hybrid model where autonomous voice and software agents take on repeatable, high-volume transactions, leaving human operators to handle exceptions, subjective edge cases, and relationship management.

To bridge the gap between pilot software and regulated environments, Kastle models its agents on typical employee onboarding. Banks configure operating policies, establish access controls, and measure digital workers on standard operational metrics. Choudhary told PYMNTS.com that institutions "can hold them to the same standard as you hold your human employees on the same metrics."

Hiring Across Engineering, Product, and Sales

To deploy the Series A capital, Kastle is hiring 18 open positions, primarily based onsite at its San Francisco office.

The technical organization is hiring at the senior and staff levels to support backend inference, voice agents, and client integrations. Active openings include a Staff AI Engineer offering between $250,000 and $350,000 annually, a Staff Engineer paying $250,000 to $320,000, and a Forward Deployed Engineer with compensation between $170,000 and $250,000.

Product and deployment operations are also scaling rapidly. The team has posted roles for an Agent Product Manager ($150,000 to $220,000) and an AI Deployments Lead ($200,000 to $250,000) to manage bank implementations. On the recruitment and commercial fronts, Kastle is hiring a Head of Talent ($200,000 to $250,000) and a Founding Growth Lead ($170,000 to $200,000).

The company also maintains entry paths through founder intern positions and a remote AI Trainer (Part-time) role compensated at $30 per hour.

Open Positions at Kastle

See all open roles at Kastle