On the back office of a small US manufacturer, the bottleneck is rarely the machines. It is the spreadsheet open on a purchasing manager's second monitor, the supplier who has not returned an email in four days, and the part that was supposed to arrive on Tuesday. Tenkara, a San Francisco seed-stage company, is building AI agents aimed squarely at that desk. The company announced a $7 million seed round led by True Ventures to expand the product to more US manufacturers [PR Newswire, 2026-03-17].
The pitch is narrow on purpose. Tenkara's agents take on what the company describes as the back-office burden for lean operations teams: finding suppliers, chasing quotes, flagging delays, and keeping production schedules honest [PR Newswire]. On its own site, the company frames the work as helping manufacturers "find better suppliers faster, prevent delays, and run a more efficient manufacturing business" [Tenkara.ai]. That is a meaningful wedge in a sector where ERP systems were largely designed before email attachments were common, and where the person doing supplier outreach is often also the person doing receiving.
The bet
Founder Benjamin Stern is explicit that the company grew out of his own frustration. "I started Tenkara after building two factories and realizing the tools didn't exist for the problems I faced," he writes on the company's about page [Tenkara.ai]. That framing matters because the AI-for-operations category is crowded with software people building for an industry they have not worked in. Stern is a 2020 Thiel Fellow with experience standing up two factories from scratch and managing production for several consumer brands [Morningstar, 2026-03-17]. He is joined by Evan Adkins on engineering and Jonah Stillman on commercialization [The Malaysian Reserve, 2026-03-31].
The product surface, based on what the company has disclosed publicly, sits between an ERP and a human buyer. Rather than replacing the system of record, Tenkara's agents act on top of it: sourcing candidate suppliers, watching for slipped commitments, and handling the routine correspondence that consumes the working hours of a small operations team [Tenkara.ai] [PR Newswire]. The two open engineering roles on the company's careers page, both software engineers focused on the platform and the React front end, suggest the team is still building out core product surface area rather than scaling a sales motion [Tenkara.ai].
Why this could be big
The tailwinds here are real. American manufacturing has become a bipartisan policy priority, and the operational reality of reshoring is that smaller US contract manufacturers, job shops, and brand-owned factories are being asked to take on volume that used to live in larger overseas plants. Those shops typically run lean, with one or two people responsible for sourcing across hundreds of SKUs. Software that can credibly handle even a slice of that workload, the supplier discovery loop, the expediting calls, the delay alerting, would compress a real cost line and, more importantly, free up scarce operator attention for the work that actually requires judgment.
True Ventures leading the round is a signal worth weighing. The firm has a long track record in early-stage technical bets and tends to underwrite founders who have lived inside the problem they are solving. Stern fits that pattern. His prior company, Nohbo, the sustainable personal care startup he founded at 16, secured a $100,000 investment from Mark Cuban on Shark Tank Season 7 [Forbes, 2016-03-15] and went on to raise a $3 million seed in 2020 [Tech Startups, 2020-03-30]. That arc, consumer hardware to factory floor to operations software, is the resume of someone who has felt the pain Tenkara is trying to address.
Funding snapshot
Tenkara seed (2026) | 7 | $M
Nohbo seed (2020) | 3 | $M
Nohbo Shark Tank (2016) | 0.1 | $M
The figures above are the disclosed funding events tied to Stern's two companies, with Tenkara's $7 million seed led by True Ventures the most recent [PR Newswire, 2026-03-17] [Tech Startups, 2020-03-30] [Forbes, 2016-03-15].
The honest counterfactual
What bears will say is that AI agents for supply chain are now a busy category, with both incumbents bolting agentic features onto existing ERP and procurement suites and a wave of seed-funded startups going after similar workflows. Selling software into small and mid-sized US manufacturers is also famously slow: buying cycles are long, IT budgets are thin, and the user is often skeptical of anything that promises to automate their judgment. What bulls will answer is that Tenkara is not trying to replace the ERP or the buyer. It is targeting the work that sits between systems, and it is being built by a founder who has run the purchasing seat himself. The company's own framing, that the agents exist so lean teams can do more, rather than so headcount can be cut, is the version of this pitch most likely to clear a plant manager's skepticism.
What to watch
The next twelve months will tell a lot. Watch for Tenkara to name design-partner customers publicly, ideally with a specific manufacturing vertical (contract electronics, consumer packaged goods co-manufacturing, and metal fabrication are the obvious early targets given Stern's background). Watch the engineering hires: a platform team build-out usually precedes a push into multi-tenant deployments. And watch for a Series A conversation in late 2026 or early 2027, which is the typical cadence for a True-led seed of this size if early usage data holds.
For a category that has spent two years generating more demos than deployments, a founder who has actually stood on a factory floor at 2 a.m. waiting for a supplier callback is a useful kind of credibility. Whether that translates into durable software revenue is the question Tenkara now has $7 million and a small team to answer.
Pulse Raman covers the systems that keep operators, clinicians, and supply chains running. Reach out with tips on deployments worth visiting in person.