Sela Raises $21M as AI Loan Agents Hit $1B in Monthly Volume

Sela, a company building AI agents for loan origination, has raised $21 million in new funding as its automation technology reaches $1 billion in loans processed every month. The milestone signals that AI-driven agents are now handling a material share of the origination workflow inside lending institutions, moving well past pilot-stage experimentation into production-level volume.

The funding will support Sela's continued build-out of its AI agent platform, which is designed to take over portions of the loan origination process that have traditionally required manual review, data entry and coordination across multiple systems. The $1 billion-a-month figure suggests the company's agents are already embedded in live lending operations rather than running as isolated proofs of concept.

Loan origination has long been one of the most document-heavy, compliance-sensitive workflows in financial services, making it an unusually demanding proving ground for AI agents. Sela's traction shows that agentic AI can now be trusted with high-stakes, multi-step processes that combine data verification, decisioning support and coordination — not just simple chat-based Q&A or customer-facing scripts.

For lenders and other regulated service providers, this points to a shift in what "automation" means: from single-task bots to agents that can own an end-to-end process and be measured in loan volume rather than ticket-deflection rates. Leaders in financial services and adjacent regulated industries should treat this as an early signal that AI agents are becoming a credible operating layer for complex back-office work, with direct implications for speed-to-decision, cost per origination and the employee experience of loan officers and processors.

— $1 billion in loans originated per month with the help of Sela's AI agents

— $21 million raised in the company's latest funding round

The headline number here isn't the funding — it's the loan volume. Raising capital is a bet on a future outcome; processing a billion dollars a month is proof the outcome is already happening.

Most coverage of AI agents still frames them as customer-facing novelties — chatbots that answer questions faster. Sela's traction shows the bigger opportunity is upstream, in the unglamorous, high-friction workflows that decide how fast a customer actually gets a loan, a claim, or an approval. The behavioral lesson for operators is simple: customers rarely complain about the interface, they complain about the wait. Any leader chasing AI ROI should stop measuring agents by conversations handled and start measuring them by decisions completed — because that's the metric borrowers, employees and regulators all actually feel.