From practice to product: turning a tool from your firm into a startup

• Speckle: TechCrunch describes its co-founders as trained architects and engineers who built the tool after struggling to exchange large 3D files. It began as a Grasshopper plug-in, raised a US$5.5m seed, and a US$12.5m Series A in 2025.
• Finch (Sweden): AEC Magazine says the idea came from the founders' own architecture practice. A floor-plan generator for Rhino went viral, practitioners asked to buy it, and the company raised a €2.5m seed in 2022.
• Monograph: co-founder Robert Yuen spent more than a decade as an architectural designer. The team ran an agency building tools for architects before turning a time-tracking tool into a practice-management product.
• Hypar: co-founder Anthony Hauck spent about 20 years in architecture and construction before product roles at Autodesk.

Architect Magazine quotes Konstru co-founder Jonatan Schumacher, who came out of engineering firm Thornton Tomasetti, on what changed: learning to market a company rather than a service. In-house R&D teams can be judged on how they help win work; a spin-off has to sell something people pay for, and faces closer scrutiny from investors.

AEC Magazine warns against products that need customising for every client: they're costly and don't scale. Monograph still treats support and onboarding as part of the product, and builds each feature with two or three customers.

In India, Section 17 of the Copyright Act, 1957 makes the employer the first owner of work an employee creates in the course of employment, and a legal note in the Legal 500 lists software and architectural drawings among such works. Someone working on a contract for service, rather than as an employee, usually keeps ownership unless they assign it.

• Read your employment contract before you build anything on the side.
• Use your own equipment and accounts, not the firm's.
• If you're still employed, get any permission to work on it in writing.
• Assign the IP to the new company when you set it up, not when you raise money. Stripe notes investors check who owns a startup's IP, and unassigned founder IP is a common problem in seed and Series A rounds.
• If the firm wants to back it, agree its stake early. Thornton Tomasetti's TTWiiN programme gives equity to the team that built a product when it becomes a company; with T2D2, the firm kept a small stake.

For tools from practice, the first customers are usually other practices. Speckle names Arup, Mott MacDonald and Aurecon among its users; Finch counts White Arkitekter and Herzog & de Meuron, and reported a 12,000-strong waitlist. Monograph's product came from one firm asking for a version of a tool built for another. T2D2 started from Thornton Tomasetti's own inspection photos.

AEC Magazine adds a caution: tools that 'makers develop for makers' can delight colleagues and still not find enough paying customers to scale.

• Stripe suggests outside money makes sense once growth potential outstrips what you can fund yourself, and you have steady revenue, customers and a plan for the money.
• Monograph bootstrapped alongside agency work and raised only when the product needed the founders full time.
• In construction tech, Nymbl reports Series A investors waiting for US$5–10m in revenue, so plan how you'll fund the stretch before it: revenue, grants, extensions or debt.
• Foundamental argues that real validation takes time, and that fast back-to-back rounds in 2020–21 hid unanswered questions about unit economics.
• In India, non-dilutive options include state grants and the Seed Fund through incubators: see our guide to grants and schemes.