BREAKING — Prodigal trained AI on 1B+ consumer-finance conversations 100+ lenders & agencies run on the platform 💰 Series A: $12M led by Accel & Menlo Ventures ✅ Up to 98% fewer compliance mistakes 🌟 4.9/5 on G2 🤖 proAgent works 24/7 across voice & digital BREAKING — Prodigal trained AI on 1B+ consumer-finance conversations 100+ lenders & agencies run on the platform 💰 Series A: $12M led by Accel & Menlo Ventures ✅ Up to 98% fewer compliance mistakes 🌟 4.9/5 on G2 🤖 proAgent works 24/7 across voice & digital
Mountain View · Est. 2018 · AI / Consumer Finance

Prodigal.

The only AI built for consumer finance - it has listened to over a billion lending and collections conversations, so your agents don't have to wing it.

AI FINTECH SAAS ENTERPRISE Y COMBINATOR
Prodigal - AI agents for loan servicing and debt collections
THE BILLION-CONVERSATION ENGINE, FRAMED AND ON THE WALL.

A debt-collection call, finally worth listening to

Somewhere right now a borrower's phone is ringing. On the other end, instead of a tired agent fumbling through a compliance script and a stack of sticky notes, there is a system that has already heard this exact conversation a million times before. It knows what to say, when to soften, and exactly which sentence a regulator would frown at. That system is Prodigal - and it is quietly rewiring one of the least glamorous corners of finance.

Consumer lending and collections is an industry built on awkward phone calls and paperwork nobody enjoys. For decades the data inside those calls - the hesitations, the promises to pay, the small print read aloud - simply evaporated the moment the line went dead. Prodigal's bet, placed back in 2018 by three IIT Bombay graduates, was that those conversations were the most valuable asset in finance, and everyone was throwing them away.

"Systems that learn, adapt, and act with human-like precision."

The company didn't start with a flashy chatbot. It started by listening. By 2019 it had analyzed 200 million interactions and found product-market fit - not by guessing what lenders needed, but by hearing it. That data became a moat, and the moat became a platform.

What a billion conversations buys you

1B+
Consumer interactions processed
100+
Companies on the platform
$14.2M
Total funding raised
15M+
Loan accounts analyzed
~91
Employees
4.9/5
Rating on G2

Eight tools, one intelligence engine

Everything connects through PIE, the Prodigal Intelligence Engine - the layer that remembers context so the agent (human or AI) never starts cold.

Autonomous Agent

proAgent

An AI agent that services and collects 24/7 across voice and digital, running operations end to end.

Core Layer

PIE

The intelligence engine trained on 1B+ interactions that carries context between every application.

Orchestration

proCollect

Omnichannel platform that runs digital collections strategy across every channel.

Scoring

proScore

Propensity-to-pay and intent scoring powered by interactions and 500+ consumer attributes.

Payments

proPay

An AI-native self-serve payment portal that lifts how many consumers pay on their own.

Documentation

proNotes

Automated, standardized call notes written in real time - now available in Spanish.

Live Help

proAssist

Real-time agent assistance with dynamic prompts during live conversations.

Quality

proInsight

QA automation, scorecards and revenue insights pulled straight from interaction analytics.

What clients actually get

Prodigal reports these as upper-bound results across deployments - directional, not guaranteed.

More agent effectiveness
30%
More digital engagement
27%
More payments
16%
Compliance errors cut
98%

Three engineers, one unglamorous industry

SG
CO-FOUNDER & CEO
SR
Sangram Raje
CO-FOUNDER & CTO
SG
Saransh Garg
CO-FOUNDER

All three met at IIT Bombay before landing in Mountain View. Gangal frames the work as building intelligence that acts "with human-like precision"; Raje runs an engineering culture he describes as rooted in empathy and real-world problem-solving. The company's five values - curious optimism, pursue excellence, bias for action, radical candor, and specificity - read less like a poster and more like a code review checklist.

From listening to acting

2018
Founded in Mountain View, backed by Y Combinator from inception.
2019
Analyzed 200M interactions and reached product-market fit; raised a ~$2M seed.
2021
$12M Series A led by Accel and Menlo Ventures; launched its AI co-pilot.
2023
Deployed fine-tuned LLMs trained on proprietary industry data.
2024
Unveiled PIE, the Prodigal Intelligence Engine, as a unified context layer.
2025
Launched proAgent and proPay, stepping fully into autonomous AI agents.

Money, backers & the fine print

Funding

  • SEED · 2019
    ~$2M from Y Combinator & Accel
  • SERIES A · JUL 2021
    $12M led by Accel & Menlo Ventures
  • TOTAL
    $14.2M across 3 rounds, 8 investors
  • EST. REVENUE
    ~$4.8M annual (third-party estimate)

Who uses it

  • AUTO FINANCE
    Lenders servicing vehicle loans
  • COLLECTIONS
    Agencies recovering consumer debt
  • LENDING & BANKING
    Consumer credit institutions
  • HEALTHCARE RCM
    Revenue cycle management teams

Four things worth knowing

Demos & interviews

Share Prodigal

Links & sources

Back to that ringing phone

Return to the borrower whose phone was ringing at the start. The call still happens - people still owe money, and money is still uncomfortable to talk about. But the conversation is different now. The script writes itself, the compliance line never gets skipped, the notes are typed before anyone hangs up, and the next person to call already knows what was said. Prodigal didn't make debt collection glamorous. It made it remember. And in an industry that spent decades letting its most valuable conversations vanish into dial tone, remembering turns out to be the whole game.