The archive · AI & Models · Product decision · 2023–2025
Momentic bets plain-English, self-maintaining tests win QA; $15M Series A
Momentic lets teams describe tests in plain English while AI maintains the scripts; 2,600 users, 200M test steps a month, $15M Series A.
Momentic
What the business is
AI-native software testing: developers describe critical user flows in plain English and an AI agent runs browser tests that evolve automatically as the app changes.
Starting capital：$3.7M seed (March 2025, YC + General Catalyst + FundersClub + AI Grant); $15M Series A led by Standard Capital with Dropbox Ventures (November 2025)
How it started
Wei-Wei Wu and Jeff An founded Momentic in late 2023 after both hit the same wall maintaining large test suites at Qualtrics and WeWork; Wu, a Node.js contributor, saw testing as 'the biggest pain point for every team I've ever worked with'.
What happened
Product launched in 2024; the $3.7M seed closed March 2025 with angels including Box CEO Aaron Levie; early customers Quora, Retool, Fiddler Labs and Runway reported 60% less debugging time and 30% faster test runs. Mobile testing arrived August 2025; the $15M Series A landed November 2025.
How it ended up
Still scaling: the Series A funds test-case management and hiring; Wu argues AI-written apps will create an endless stream of new products that all need verification.
Background
Momentic started from a frustration both founders shared: test automation at scale is a treadmill. Wei-Wei Wu and Jeff An, who built developer tooling at Qualtrics and WeWork, kept seeing teams write Selenium-style scripts that break every time the UI changes, then spend more time repairing tests than shipping features. Their answer, founded in late 2023, was to turn testing into an AI task: describe a critical user flow in plain English and an agent writes, runs and fixes the browser test itself.
The product sells the maintenance story, not the framework story. Instead of fine-grained control over selectors and assertions, Momentic keeps the description in natural language so the test script can automatically evolve as the application changes, 'as long as the intent stays the same,' Wu told SiliconANGLE in March 2025. Early adopters Quora, Retool, Fiddler Labs and Runway reported a 60% cut in debugging time and 30% faster test execution.
Money followed the narrative: a $3.7 million seed in March 2025 from Y Combinator, General Catalyst, FundersClub and AI Grant — with Box CEO Aaron Levie among the angels — then a $15 million Series A in November 2025 led by Standard Capital with Dropbox Ventures. By then Momentic counted about 2,600 users across Notion, Xero, Bilt, Webflow and Retool, and claimed more than 200 million test steps automated in a single month.
What has to be true
- It attacked the cost that actually hurts: script upkeep. Framework players compete on features; Momentic competed on making test maintenance disappear, the pain engineering teams feel weekly.
- Plain-English intent is a durable moat against foundation models, because the model vendors sell agentic testing tutorials, not a product that keeps tests in sync with a living codebase.
- The dosing was right: nontechnical users get automatic test execution, technical users get a low-code editor to fine-tune, so sales didn't require convincing QA teams to abandon rigor.
- Anchoring on named logos with measurable outcomes (60%/30% improvements) gave the seed-to-A round story credibility that purely technical differentiators wouldn't have.
What can be applied
Sell the maintenance, not the build: the wedge that worked was that tests stop rotting — describe intent once and the script evolves with the product, converting a hated chore into a set-and-forget.
Aftermath
As of the November 2025 announcement, Momentic is using the Series A to build more sophisticated test-case management and hire engineers. The biggest strategic risk it names is the model layer itself: OpenAI and Anthropic publish tutorials on agentic testing, and as foundation models grow stronger at computer use, the headroom for a dedicated testing SaaS may narrow. Wu's counter-bet is that the wave of AI-generated applications creates so many new products that demand for independent verification grows even faster.
Sources
- Momentic raises $15M to automate software testing
- Momentic raises $3.7M to redefine software testing with AI
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card