Freehand, a San Francisco startup building autonomous AI agents for enterprise procurement, has raised $75 million to expand its technology that handles invoice auditing, payment processing, and vendor negotiations without human intervention. Battery Ventures and NewRoad Capital Partners co-led the round, with participation from PSP Growth (the venture firm tied to former Commerce Secretary Penny Pritzker), Nexus Venture Partners, and others.

The company emerged from stealth with live deployments at Meta, Unilever, Johnson & Johnson, Pfizer, Dunkin', and Cardinal Health. Early customers report recovering 5 to 10 percent of spending in complex categories, executing processes five to seven times faster, and cutting procure-to-pay timelines by more than 70 percent.
What Freehand's AI agents actually do
Freehand's agents operate inside existing ERP systems. They review contracts, negotiate supplier terms, flag billing discrepancies, process payments, manage vendor relationships, and reconcile data. The company positions this as a direct replacement for legacy software stacks combined with large outsourced teams.
At the core sits what Freehand calls its Category Context Graph. The system merges unstructured data from documents and communications with structured enterprise records, aiming to give agents the contextual knowledge of experienced supply chain professionals. Every action generates an audit trail, and the company claims the graph improves accuracy over time through continuous enrichment.
Co-founder Abhijeet Manohar drew a line between Freehand's approach and conventional copilot tools. "We're shifting from tools that merely assist users to systems that themselves function as the operator," he said, arguing that deep contextual understanding separates genuine agents from suggestion engines.
The $348 billion labor gap
CEO Nitin Jayakrishnan, who previously founded and sold logistics software company Pando, framed the opportunity in stark terms. Enterprises spend roughly $16 billion annually on supply chain software while allocating an additional $348 billion on personnel to perform tasks that software still cannot handle. Freehand aims to close that gap with agents that decide, act, and accept responsibility for outcomes.
Invoice auditing and payment management alone can cost large organizations tens of millions per year. The startup targets these functions first, with plans to expand into direct materials and maintenance, repair, and operations spending.
Why investors wrote the check
Battery Ventures general partner Dharmesh Thakker, who joins Freehand's board, cited the company's vertical focus and traction with major enterprises. He contrasted its decision-making agents with basic copilot products. NewRoad's Gregoire Lehmann pointed to measurable ROI: reduced overpayments, lower operating costs, stronger audit coverage, and automation of manual workflows.
“Operational efficiency at major firms is critical for broader industrial competitiveness. Freehand converts advanced AI into accountable productivity gains.”
— Penny Pritzker, PSP Growth
The funding arrives as tariffs, tax changes, and shifting immigration policies pressure global supply chains. Companies responding to these pressures are reportedly shifting internal staff toward strategic work and scaling back business process outsourcing arrangements.
Logicity's Take
Freehand's 70% reduction in procure-to-pay timelines is the number to watch. If that holds at scale, it undercuts the economics of outsourced BPO providers who charge per transaction or FTE. The real test: whether enterprise legal and compliance teams will let autonomous agents negotiate supplier contracts and authorize payments without human sign-off. The audit trail feature suggests Freehand anticipated that objection. Finance teams evaluating this space should also look at Coupa, SAP Ariba, and emerging AI-native competitors like Zip to benchmark capabilities and pricing.
What the announcement leaves open
Freehand did not disclose pricing, contract terms, or the size of its customer deployments beyond naming the logos. The company also did not specify what percentage of procure-to-pay tasks its agents handle end-to-end versus those still requiring human approval.
The 5 to 10 percent spend recovery figure sounds significant, but it depends on the category. Complex indirect spend categories with high error rates and fragmented vendor bases will show better returns than commoditized direct materials. Enterprises evaluating Freehand will want to see category-specific benchmarks and understand where the product's autonomy ends.
Another enterprise AI startup raising significant capital to automate human-intensive business functions
Need Help Implementing This?
If your finance or procurement team is evaluating AI automation for spend management, reach out to Logicity for vendor comparisons and implementation guidance.
Source: Crowdfund Insider
Manaal Khan
Tech & Innovation Writer
Produced with AI assistance and reviewed by the Logicity editorial team. Learn more in our Editorial Policy.






