Atira raises $17.5M after landing industrial customers without a sales team

Why it matters

Atira found a practical enterprise AI wedge in the paperwork that precedes industrial production. Its founder-led traction is credible; the seed round will test whether that sales motion and product depth can scale internationally.

Atira co-founders Florian Diegruber and August DuMont Schuette have raised $17.5M for an AI system that helps manufacturers turn dense technical requests into configured, priced bids.

The financing, reported September 3rd by FinSMEs and announced by Atira, includes a $15M seed round led by Accel and a previously unannounced $2.5M pre-seed. UVC Partners, Fortino Ventures and Booom backed Atira across the financing, while the individual investors include Whirlpool Chairman and CEO Marc Bitzer and Celonis co-founder and co-CEO Bastian Nominacher.

The round follows an unusually founder-heavy first year of selling. Accel said Atira signed more than a dozen customers with no commercial employees, with Diegruber and DuMont Schuette closing the accounts themselves and customer referrals supplying much of the pipeline. UVC Partners and Atira put the total at more than 15 signed customers. Fortune reports that roughly 15 customers were using the system in production, including ABB E-Mobility, Chiron Group, Rema Tip Top, Robel, Theegarten-Pactec and Fette Compacting.

That record explains what Accel is funding: a pair of technical founders who found a neglected enterprise workflow, sold the product themselves and reached production deployments before building a conventional go-to-market organization.

An unsexy wedge into manufacturing

Diegruber previously led commercial work for major industrial accounts at Palantir in German-speaking Europe. DuMont Schuette worked as a software engineer at Google, where Accel says he shipped large language model systems handling hundreds of millions of queries per day. His earlier research included work on generative modeling and privacy-preserving medical data at the Max Planck Institute for Intelligent Systems and ETH Zurich.

The founders left Palantir and Google in 2024 with a broad thesis around the reindustrialization of Europe and the United States. They eventually settled on sales engineering, the work required to translate a customer's technical request for quotation, or RFQ, into a proposal that the manufacturer can actually build and price.

Quoting software "sounded very unsexy back then, which is always a good place to start a venture," Diegruber told Fortune.

The choice gave Atira a narrow entry point into large manufacturers. An RFQ for a rail-maintenance vehicle, charging system or custom production line can contain hundreds or thousands of pages of drawings, standards, contractual requirements and product specifications. Sales engineers must identify what the manufacturer can supply, locate exceptions, consult specialist departments and assemble technical and commercial documents. The process can run for weeks or months before a price reaches the buyer.

Factories have spent decades automating production. The work that creates the order still moves through PDFs, spreadsheets, inboxes and the memories of experienced engineers.

What Atira actually automates

Atira connects to a manufacturer's existing CRM, ERP and configure-price-quote systems. Its agents extract requirements from incoming documents, flag risks and unsupported specifications, help configure a feasible product, and draft the technical and commercial material needed for a bid.

The product also routes questions to human specialists. Diegruber told Fortune that Atira can contact the appropriate employee through Microsoft Teams when a requirement cannot be resolved from existing documents. Users see the supporting sources and confirm agent outputs, an important control when an incorrect specification can distort pricing, create a compliance problem or commit a manufacturer to an impossible configuration.

Atira describes the product as a "commercial brain" between CRM and ERP. That positioning reaches beyond document processing. Each reviewed proposal can add customer requirements, accepted configurations, pricing decisions and expert answers to a reusable knowledge base.

The immediate performance figures come from Atira and its customers. Atira says customers have cut ask-to-bid time by as much as 80%. In an Atira case study, railway-equipment manufacturer Robel said the platform saves 95 hours of sales and engineering work per RFQ. The case study says 18 specialist departments review tender materials and Robel uses more than 20 internally developed Atira agents in its tender-response process.

Chiron Group, an early design partner, has deployed Atira to more than 70 users and reports an 80% reduction in ask-to-bid time. Atira says ABB E-Mobility has deployed the platform globally. These figures have not been independently audited, but the production deployments carry more weight than pilot counts in a market crowded with agent demonstrations.

Founder-led sales, then a commercial organization

Diegruber told Fortune that Atira launched commercially in November 2025 and had about 15 employees by September 2026, up from four at the start of the year. Nine additional hires were expected to join, and the seed capital will primarily fund engineering and Atira's first commercial employees.

UVC Partners, which first backed Atira in the $2.5M pre-seed and increased its commitment in the seed, said the financing will also support international expansion and adjacent products such as pricing intelligence, aftersales and supplier coordination. Atira already has a US customer in a pilot, according to Diegruber.

Hiring a sales organization changes the operating test. Founder-led selling lets Diegruber and DuMont Schuette translate customer complaints directly into product decisions. A larger commercial team must reproduce that technical credibility across manufacturers with different product catalogs, internal systems and approval processes. Fast initial deployment matters little if every account later demands extensive custom engineering.

Diegruber told Fortune that customers can begin using the product within a day or two and cancel after a month. He also said none had canceled. Atira has not published revenue, pricing or gross-margin figures, leaving the cost of supporting those deployments outside view.

AI moves in front of CPQ

Atira is entering a category that established software vendors and younger startups are already expanding around. The common target is the unstructured material that arrives before a conventional CPQ system can calculate a valid configuration and price.

In May, manufacturing CPQ vendor Tacton launched an AI modeling assistant that converts technical documents and spreadsheets into structured product models for human review. In June, Threekit released an AI sales agent that interprets notes, drawings, RFPs and other buyer inputs before recommending valid configurations. Industrial equipment quoting startup Corbel announced a $6.7M seed round for a platform combining equipment configuration, quoting and embedded financing.

Atira's bet is that the winner will need to coordinate the entire bid process, including document review, specialist input, product configuration and proposal generation. That would put Atira above existing systems as the layer through which commercial decisions move.

The $17.5M gives Diegruber and DuMont Schuette room to test that claim outside their early founder-sold accounts. Their first customers have already shown that manufacturers will buy AI for an overlooked, paperwork-heavy workflow. The next job is making those deployments repeatable without losing the industrial knowledge that won them in the first place.