HomeCompaniesLighthouz AI

AI Agents for AP and AR in Logistics

Lighthouz AI's AI agents solve the biggest problems in logistics - AR and AP! 3PLs, brokers, factors, and carriers lose billions every year to slow invoicing, manual invoice audits, and billing disputes. Manual AP & AR processes create errors, disputes, cash flow bottlenecks, late fees —hurting everyone in the logistics supply chain. 🚛 The Problem: - Factors: Auditing and verifying carrier invoices, timely invoicing - 3PLs & Brokers: Verifying carrier invoices, creating own invoices, and handling disputes and chargebacks. - Carriers: Timely invoicing, Submitting correct and complete paperwork, Dealing with late payments, incorrect deductions, and manual paperwork. - Shippers: Struggling with invoice error identification and resolution, delaying payments. 📉 The Impact: The U.S. freight industry processes over $1.2 trillion in transactions annually, yet 25% of freight invoices contain errors, leading to payment disputes and delays. 🔹 Our Solution is an AI agent that automates invoice audits, rate verifications, invoicing, and collections in the logistics industry. With Lighthouz AI agents, businesses 10x their AR and AP processes.
Active Founders
Srijan Kumar
Srijan Kumar
CEO & Co-founder
I'm an award-winning AI professor-turned-entrepreneur. Over the last 14 years, my AI systems have been used at Meta, Wikipedia, The Home Depot, and Flipkart. I have been a Georgia Tech professor and worked at Google, Stanford, CMU, and UMD, and IIT. I pioneered dynamic graph neural nets. I have written 60+ AI publications w 6000+ citations. Awards won: NSF CAREER, Forbes 30u30, Kavli Fellow. I'm building AI agents to do AP and AR in logistics, powered by dynamic knowledge graphs.
Sonali Pattnaik
Sonali Pattnaik
CTO & Co-founder
Sonali is the CTO and Co-founder of Lighthouz. Previously, she has led AI teams at enterprises such as Progressive, American Family, and Halliburton, where she spearheaded innovative AI solutions. With experience in building and deploying AI technologies, she holds multiple US patents and papers. Working across multiple industries, Sonali has gained expertise in tackling business problems from inception to full implementation using AI solutions.
Company Launches
Lighthouz AI: AI to automate AP and AR for freight brokers and 3PLs
See original launch post

https://www.youtube.com/watch?v=SNdu7kqN0MY

Freight brokers and 3PLs still run AP and AR manually.

Accounting teams spend days processing carrier bills, chasing missing PODs, delaying customer invoicing, and collections efforts. Cash flow stagnates. Before Lighthouz, teams typically take 3–5 days to approve carrier bills and send customer invoices — and still miss errors like duplicate payments and invalid accessorials.

At Lighthouz, we’ve built freight-native AI agents to automate AP and AR for brokers and 3PLs. With Lighthouz, brokers are invoicing customers within 24 hours of delivery and reduce days-to-get-paid by 5 days.

Who we are

We’re Srijan and Sonali, the founders of Lighthouz AI. We’re part of YC S24.

We met a decade ago during undergrad at IIT, have been building together ever since, and have been married for 5+ years. Srijan and Sonali have 15+ years of combined experience in AI and supply chain, building AI systems for companies like Halliburton, The Home Depot, Google, and more.

  • Srijan an award-winning AI scientist with 60+ publications and 6,000+ citations.  He’s been a CS professor at Georgia Tech, an ex-research scientist at Google, completed his postdoc at Stanford, and earned his PhD from the University of Maryland.
  • Sonali is a former Lead AI Engineer at Halliburton, Progressive Insurance, and American Family, specializing in unstructured data and building production-grade AI systems from scratch. She earned her MS from the University of Washington.

We’ve spent our careers building and deploying AI for complex, messy real-world data — which is exactly what freight finance looks like.

What Lighthouz AI does

Lighthouz handles the entire freight billing workflow:

✅ Ingests carrier bills, PODs, BOLs, lumpers, and invoices from email and uploads

✅ Extracts and structures all freight data — rates, accessorials, remit-to, signatures, handwritten notes

✅ Verifies carrier bills against rate confirmations / quotes

✅ Detects exceptions like duplicates, overbilling, missing PODs, wrong factors, and invalid accessorials

✅ Auto-approves clean bills and routes only true exceptions

✅ Approves and submits customer invoices within 24 hours of delivery

✅ Sends automated collections reminders to customers pre- and post-due date

✅ Creates a work queue for human collectors

✅ Posts clean, validated data directly into the TMS with full audit trails

Why this matters

Manual AP/AR quietly kills cash flow in freight. Before Lighthouz, teams typically take 3–5 days to approve carrier bills and send customer invoices — and still miss errors like duplicate payments and invalid accessorials.

With Lighthouz:

  • Processing time drops from 3–5 days → <24 hours
  • Days to get payment from customers reduces by 5 days
  • One AP specialist can process 4–5x more volume
  • Errors and rework drop dramatically
  • Teams scale without adding headcount

This isn’t generic automation. Lighthouz is freight-native, built for multi-stop loads, multi-document PDFs, missing pages, rotated scans, handwritten notes, and factoring response workflows.

Where we are

We are already working with leading freight brokers and 3PLs in the US and Canada.

Ask ❤️

If you know freight brokers or 3PLs, please introduce us!

Jobs at Lighthouz AI
Remote (US)
$150K
3+ years
IN / Remote (IN)
₹3M INR
1+ years
IN / Remote (IN)
₹10 - ₹10M INR
6+ years
IN / Remote (IN)
₹4M INR
3+ years
Lighthouz AI
Founded:2023
Batch:Summer 2024
Team Size:4
Status:
Active
Location:San Francisco
Primary Partner:Tom Blomfield