Loop, a San Francisco-based AI company, launched its Logistics Data Platform (LDP) to solve pervasive supply chain data fragmentation. Backed by a $95 million Series C round led by Valor Equity Partners, the platform features DUX 2.0—a domain-specific LLM extracting 200+ data points per shipment—and autonomous Exception Agents that resolve delivery anomalies in hours instead of weeks. CEO Matt McKinney emphasized that agentic AI requires pristine, supply-chain-native data, citing lessons from Uber Freight. The platform aims to transform logistics from human-operated execution systems into AI-driven systems of action amid rising tariffs, energy costs, and nearshoring pressures.
On Monday, San Francisco-based AI company Loop unveiled an expansive suite of capabilities that form the foundation of its Logistics Data Platform (LDP). The launch is aimed at tackling systemic logistics data challenges plaguing the supply chain.
The $95 Million Series C Backing Expansion
The launch follows a recent $95 million Series C funding round led by Valor Equity Partners and the Valor Atreides AI Fund, with participation from 8VC, Founders Fund, Index Ventures, J.P. Morgan Growth Equity Partners, and Tao Capital Partners.
“Automation and AI are only as powerful as the data foundation they operate on,” said Matt McKinney, CEO and co-founder of Loop. “With this launch, Loop fundamentally shifts from a platform that helps humans work faster to an autonomous system that executes complex logistics strategies on their behalf. We are delivering the pristine data foundation required to bring agentic AI to the supply chain.”
The Bad Data Problem That Plagues Supply Chains
For decades, logistics teams have grappled with fragmented, unstructured data trapped in emails, physical documents and disconnected systems. This caused a cascade of silo-related issues. Transportation struggles to optimize costs. Finance lacks visibility into true landed costs. Customer care operates reactively. Transformation teams watch AI pilots fail because the underlying data simply isn’t there.
McKinney traces the problem back to his days at Uber Freight. “When we were on the freight product at Uber Freight, we found an industry-wide problem: invoice reconciliation,” McKinney said. “As we got into that and realized what was causing all the issues and exceptions, we discovered it was bad data.”
The breakthrough came faster than anyone expected. Large language models (LLMs) that McKinney anticipated reaching maturity in 2030 delivered in 2025, making Loop’s mission possible: unlocking value trapped in operations to power the physical economy.
DUX 2.0: The Data Engine Powering Agent-Ready Intelligence
At the platform’s core sits DUX 2.0, Loop’s next-generation domain-specific language model designed to extract and normalize data from PDFs, emails, spreadsheets and enterprise resource planning (ERP) systems. The engine collects over 200 data points per shipment and expands into customs, tariff document review and purchase order matching.
The key part is tailoring the solution to speak the specific language of the supply chain, something that a general LLM may find difficult to understand the nuances. The other issue is that it’s costly to make mistakes.
“You can’t take an off-the-shelf general horizontal LLM and apply it to supply chain,” McKinney said. “Supply chain runs on its own language, and those things hallucinate. And guess what? If you lose trust with your vendor or your customer, you’re done.”
The Exception Agent: From Weeks to Minutes
Loop’s Exception Agent operates as an autonomous “AI teammate” that manages touchless exception handling — initiating disputes with carriers, resolving payment queries and ensuring accurate invoice payments. McKinney described how “swarm agents” work together, with process validation agents checking each other’s work to maintain accuracy.
“The exception is the rule in supply chain,” McKinney said. “If you don’t have an exception agent that can autonomously resolve these exceptions and initiate disputes, you’re not gonna be able to unlock your team’s time.”
The results speak for themselves. One large retailer had 200 customer packages inadvertently delivered to pickup locations instead of residences. “An agent detected this — literally the exception agent took this in hours and immediately notified the carrier that there was an issue and they should go pick up these packages,” McKinney said. “So the next day, the carrier was able to pick up those packages and get them to the customers’ homes on time.” Previously, resolution would have required customer complaints filtering through care teams, manual investigation and weeks of back-and-forth.
Loop Intelligence and the Strategic Layer
The platform’s Loop Intelligence feature transforms company data into strategic power through an AI Assistant that accepts natural-language prompts for network optimization and financial clarity. McKinney positioned the platform as a layer that orchestrates across existing execution systems.
“Think of those as execution systems that are really databases, and they are a tool that humans use,” McKinney said. “Agents actually do all the work. So those are the system of records. Agents are the system of action.”
Looking Ahead
The $95 million investment arrives as supply chains face volatile operating conditions — rising energy costs, tariffs and nearshoring pressures reversing 50 years of globalization trends. “Supply chain is one of the largest contributors to GDP, and it’s one of the industries that hasn’t really been touched by technology in the same way because there just wasn’t the same level of investment other industries had,” McKinney said.
Source: FreightWaves
Compiled from international media by the SCI.AI editorial team.
