Momentic raises $15M to automate software testing

Product demos get each the attention, but bundle improvement much often involves things similar debugging, prime assurance and testing. It’s the dull but captious enactment that keeps bundle moving the mode it should, and arsenic developers look to automate much of their workloads, it’s progressively being done by AI.

On Monday, the AI investigating startup Momentic said it had raised $15 cardinal successful a Series A circular led by Standard Capital, with information from Dropbox Ventures. Existing investors astatine Y Combinator, FCVC, Transpose Platform and Karman Ventures besides participated successful the round. The caller backing builds connected a $3.7 cardinal effect round, which the institution announced successful March.

Momentic makes tools for bundle investigating and verification, a niche presently occupied by open-source frameworks similar Playwright and Selenium. Those tools connection complex, fine-grained controls, but Momentic is counting connected AI to marque the process elemental and effective.

“We assistance our customers marque definite their merchandise works,” co-founder Wei-Wei Wu said. “They tin picture their captious idiosyncratic flows successful plain English and our AI volition automate it.”

Wu and his co-founder Jeff An some person backgrounds successful developer tooling astatine companies similar Qualtrics and WeWork. (Wu is peculiarly arrogant of his contributions to the open-source Node.js.)

The biggest changeless for each those companies, arsenic Wu saw it, was the occupation of verifying code. “Testing has been the biggest symptom constituent for each squad I’ve ever worked with,” Wu told TechCrunch.

Momentic’s AI-driven attack has already won implicit a fig of clients. The institution presently boasts 2,600 users crossed its lawsuit base, which includes companies similar Notion, Xero, Bilt, Webflow and Retool. Wu was coy astir gross and profitability figures, but says the merchandise has shown capable maturation to person investors.

Notably, automating tests makes them overmuch easier to execute astatine scale, and has the imaginable to thrust up the full measurement to levels that would antecedently person been impossible. Wu estimates that, successful the past month, the institution automated much than 200 cardinal trial steps.

The company’s biggest rival whitethorn beryllium the instauration models themselves. Both OpenAI and Anthropic connection tutorials connected agentic testing, gathering connected their models’ increasing machine usage capabilities. As those models turn much sophisticated, the accidental for endeavor SaaS companies similar Momentic could narrow.

For now, Wu is focused connected fleshing retired his merchandise with the caller funding. The institution launched enactment for mobile environments successful August, and is hoping to physique much blase test-case absorption erstwhile it has a fewer much engineers connected board.

As Wu sees it, the emergence of automated coding volition nutrient a batch of caller apps — and a batch much demands for products similar his. “All of these apps request testing,” helium said. “They attraction astir quality, and we’re going to supply it for them.”