HomeCompaniesAccend
Accend

AI agents for commercial underwriting at banks and fintechs

Accend helps commercial banks and fintechs underwrite business customers faster, win more deals and increase revenue. We provide an end-to-end platform to automate document intake, financial statement / tax spreading, cash flow modeling, credit memos and portfolio management, with audit-ready AI and human-in-the-loop accuracy. Companies like Corpay, Brex, Rippling, Rho, Column, Settle, and many others have reduced application processing time by 80% with Accend, significantly enhancing their credit underwriting efficiency Before starting Accend, the founders led Product & Engineering on the Risk team at Brex as well as at Deutsche Bank and Credit Suisse so they have firsthand experienced issues with operational efficiency in credit/underwriting teams. Accend has raised a $3.2M seed round from Y Combinator, Adverb Ventures, General Catalyst and 645 Ventures.
Active Founders
Pranjal Daga
Pranjal Daga
Founder
Currently, Cofounder of Accend. Previously, I led Product for Risk at Brex & saved $20M in fraud losses. Before that, I dropped out of my PhD in AI/ML at Purdue University to help set up Cisco Innovation Labs where I grew the team from 1 to 35 in 4 years.
Yutong Pei
Yutong Pei
Founder
Yutong Pei was most recently the Engineering Manager on the Fraud Risk team at Brex where he built the Disputes service, worked on fraud controls by managing engineers and data scientists. Previously, he has led Fraud Engineering and Security for IoTeX, Uber and Amazon.
Joseph Zhou
Joseph Zhou
Founder
I'm one of the Co-Founders of Accend. Previously, I served as the CFO of a mobile gaming startup with a 50M ARR. I established the company’s operation in Europe and scaled the team from 0 to 80 members.
Company Launches
Accend For Banking
See original launch post
Introduction

Accend for Banking is AI infrastructure that helps banks turn financial documents into structured financial models and credit insights — from hours to minutes.

Built for commercial banks and CRE lenders, it sits inside existing credit workflows and removes the most manual parts of underwriting, so banks can move faster and win more deals.

https://youtu.be/ZZZYw-Q0XLQ


The Problem

Credit underwriting in banks is slow because it still relies on legacy systems and outdated tools that haven’t evolved in decades.

Across most banks today:

  • Financial documents are collected across emails, portals, and shared drives
  • Borrowers upload tax returns, financials, and supporting documents across fragmented systems
  • Data is manually extracted and spread into Excel
  • Financial models are rebuilt repeatedly for each case
  • Credit memos are written from scratch every time
  • Monitoring and covenant tracking happen separately

These workflows are:

  • Fragmented across systems
  • Manual and time-intensive
  • Difficult to scale without adding headcount
  • Prone to inconsistencies and errors

Even as lending demand has increased, the infrastructure supporting underwriting has remained largely unchanged for decades.


What This Means for Banks

The impact of these workflows goes beyond operations — it directly affects people, speed, and business outcomes.

  • Analysts spend hours on data preparation instead of credit decisions
  • Turnaround times slow down lending and delay decisions
  • Banks lose deals to faster competitors
  • Productivity is lost across the workflow due to repeated manual effort
  • Scaling requires adding headcount instead of improving systems

At a time when speed matters more than ever, these constraints make it harder for banks to compete.


How Accend Helps

Accend brings these fragmented steps into a single, structured workflow.

With Accend:

  • Borrowers upload tax and financial documents through a single secure link or connect their accounting systems
  • Automated requests and reminders keep document collection moving
  • Financial documents are parsed into structured, audit-ready data in minutes
  • Detailed schedules and supporting statements are captured with high accuracy
  • Property-level data (e.g., Schedule E, Form 8825) is broken down with full visibility
  • Cash flow models are generated instantly and mapped consistently
  • Credit insights and memos are built on top of structured data

This means:

  • Analysts spend time on credit decisions, not data preparation
  • Turnaround times shrink from hours to minutes
  • Teams can handle more volume without increasing headcount
  • Outputs remain consistent across deals

How Accend Works

Accend integrates into your existing credit workflow and processes each case end-to-end:

1. Document Intake

Send a secure link to borrowers to upload tax and financial documents, or connect directly to accounting systems. Automated requests and reminders ensure completeness.

2. Data Extraction & Structuring

Accend parses full tax packages in minutes — including 1040s, 1065s, 1120s, 1120-S, and supporting schedules — and converts them into structured, audit-ready data with source traceability.

3. Detailed Financial Spreading

Financials are standardized across income statements, balance sheets, and cash flows. Supporting statements and add-backs are surfaced automatically.

4. Modeling & Analysis

Data maps instantly into cash flow models. Teams can drill into sources, override values, run ratios, and leave notes — with every change tracked.

5. Configurable Financial Models

Models are configured to your bank’s policies — personal, business, global — ensuring consistency across teams while supporting complex analysis.

6. Monitoring & Covenant Tracking

Once underwriting is complete, covenants are tracked automatically with scheduled tests, due dates, and alerts.


Get Started

If you work in credit, underwriting, risk, or lending operations, Accend helps you modernize your workflow without replacing your systems.

Schedule a call with Pranjal, our co-founder, to learn more.

Or explore Accend for Banking → https://www.withaccend.com/banking

Previous Launches
Accend helps banks and fintechs collect, spread and analyze financial data for faster and more accurate underwriting using human-in-the-loop AI agents
Human-in-the-loop AI Agents for Financial Statement Analysis
Improve Ops Efficiency & Quality + Excel in Compliance with Accend
Jobs at Accend
San Francisco, CA, US / Remote (San Francisco, CA, US)
$120K - $180K
1+ years
New York, NY, US
$65K - $120K
1+ years
Accend
Founded:2023
Batch:Summer 2023
Team Size:18
Status:
Active
Primary Partner:Aaron Epstein