Dominic Toselli was the Head of Worldwide Logistics Procurement at Apple, where he watched teams drown in quote requests, contract redlining, and supplier negotiations. Now he's selling a solution: AI agents that promise to do the grunt work while procurement professionals focus on strategy.
His company, Aron, disclosed a $6 million seed round led by Storm Ventures, with total funding reaching $8 million when combined with an earlier $2 million pre-seed led by Menlo Ventures. Japan Airlines Ventures and Plug and Play participated in the pre-seed. The San Francisco startup launched publicly on September 14, 2026, after rebranding from its original name, NegotiateAI.
Aron claims its platform automates procurement tasks without requiring extensive integrations. The platform uses AI agents to chase supplier quotes over email, redline contract renewals against company playbooks, and flag discrepancies between invoices and contracts. Teams can feed work into the system by forwarding emails, dropping messages in Microsoft Teams, or pointing it toward SharePoint folders. The company claims an average 21% savings on initial requests for quotation, with no software integrations required upfront.
"Teams get their time back, and the CFO gets a number they can trust with fast ROI," Toselli said when announcing the funding.
The product requires human sign-off before executing tasks, but handles the legwork autonomously. Behind the interface sits what Aron describes as a proprietary knowledge graph that connects spend data, contract terms, and project details across email, Slack, Teams, and phone channels. It's an attempt to create what Ryan Floyd, founding managing director at Storm Ventures, calls "the first agentic AI Chief of Staff purpose-built for procurement."
TransPak, a packaging company, deployed Aron for maintenance, repair, and operations procurement and saw returns within weeks. "We started in MRO, saw ROI in the first month and deployed into our other categories," said Scott Eisenberg, the company's vice president of global supply chain. Aron says it serves customers ranging from mid-market firms to Fortune 10 enterprises, though it hasn't disclosed specific names beyond TransPak.
The company enters a procurement automation market that has attracted steady venture interest despite the broader funding slowdown. Pactum raised $20 million in November 2023 for autonomous negotiation agents. Arkestro pulled in $36 million in May 2025 for what it calls predictive procurement orchestration. Newer entrants like Zip, Levelpath, ORO Labs, and Globality are targeting specific procurement workflows with AI, carving out niches in a space long dominated by enterprise heavyweights such as Coupa, SAP Ariba, Ivalua, and Jaggaer.
Steve Sloane, a partner at Menlo Ventures, framed the opportunity in stark terms: "Procurement is one of the largest and least digitized functions in industrial supply chains."
Toselli, who holds graduate degrees in engineering from UC Berkeley and the University of Waterloo, plans to channel the new capital into hiring engineers and expanding the product. The company had between 11 and 50 employees listed on LinkedIn as of mid-September. Aron operates formally as NegotiateAI Inc., with offices at 548 Market Street in San Francisco.
Whether AI agents can truly replace the human judgment required in high-stakes supplier negotiations remains an open question. For now, Aron is betting that procurement teams care less about full automation and more about reclaiming hours lost to administrative tedium. The company scheduled appearances at several industry conferences, including Tomorrow's Procurement Leader, Catalyst, ProcureCon East, and DPW Amsterdam, to make its case directly to potential buyers.
The test will be whether enterprises trust algorithms to handle negotiations that can swing millions of dollars, or if they prefer humans to stay firmly in the driver's seat with AI playing sidekick. Aron's approach splits the difference, keeping approval workflows in place while automating the back-and-forth that typically consumes days or weeks. That middle path may prove either pragmatic or limiting, depending on how quickly buyers warm to letting software negotiate on their behalf.
