Analysis
Freehand, which builds AI agents to manage Fortune 500 supply-chain spend and back-office procurement, raised a $75 million Series B co-led by Battery Ventures and NewRoad Capital Partners, with PSP Growth and Nexus Venture Partners also participating. The company's agents are already live managing supply-chain operations for Meta, Unilever, Johnson & Johnson, Pfizer, Dunkin' and Cardinal Health.
Founders Nitin Jayakrishnan and Abhijeet Manohar are repeat operators in this exact category -- they previously built and sold Pando, a transportation-management and procure-to-pay system used by large enterprise logistics teams, before founding Freehand in February 2024. That domain credibility likely explains how quickly Freehand landed marquee logos like Meta and Pfizer, customers that rarely move fast on unproven procurement software from first-time founders.
“Those are the kind of concrete, board-reportable ROI numbers that separate durable AI-agent enterprise deployments from pilot projects that never convert to renewal.”
The reported results are strong by enterprise-software standards: customers recovering 5-10% of supply-chain spend, workflows completing 5-7x faster, and procure-to-pay cycle times cut by more than 70%. Those are the kind of concrete, board-reportable ROI numbers that separate durable AI-agent enterprise deployments from pilot projects that never convert to renewal.
The round size itself is a signal: $75 million is roughly 3x Freehand's $25 million Series A from March 2024, a step-up ratio consistent with how fast AI-native enterprise software companies are compounding round-to-round when usage and retention data backs up the pitch. Freehand competes in a crowded procurement and supply-chain AI category against incumbents like Coupa and SAP Ariba, as well as newer AI-native entrants, but its early logo list suggests real differentiation on execution speed.
What to watch: whether Freehand's Fortune 500 logo list keeps expanding at the current pace, and whether the 5-10% spend-recovery figure holds up as a durable benchmark once more customers are a full year into deployment rather than early pilots.
