Skip to main content
← sales

Meticulous Replays User Workflows to Verify AI-Generated Frontend Code

By Zero G Talent•

Software testing tools traditionally require developers to script every scenario, assert outcomes upfront, and maintain fragile test files as applications change. Meticulous takes a different approach by building automated end-to-end testing systems that run without hand-written scripts. The company's tool injects a single line of JavaScript into non-production environments to observe real user workflows, stub network responses, and replay sessions against pull requests in an augmented, deterministic browser engine. Visual and behavioral diffs surface differences before code reaches production.

CEO Gabriel Spencer-Harper co-founded the business in 2021 alongside his brother Quentin Spencer-Harper, who serves as CTO after leading frontend engineering teams at Palantir. Gabriel Spencer-Harper previously worked as a software engineer at Dropbox and Opendoor, where he observed engineering teams spending hours manually clicking through workflows to catch bugs while discarding recorded user sessions. That experience formed the basis for Meticulous, which went through Y Combinator's S21 batch.

According to yespress.io, the company raised a Series A round led by Chemistry in July 2026, with participation from Menlo Ventures, Coatue, Y Combinator, Soma Capital, Base Case Capital, and individual software leaders. Its customers include engineering organizations at Notion, Dropbox, Wiz, ElevenLabs, Discord, LaunchDarkly, CoreWeave, and Wealthsimple.

Deterministic Replay and Code Verification

Speaking at an AI Engineer conference, Gabriel Spencer-Harper explained that AI coding agents have shifted the primary engineering bottleneck from writing code to reviewing and verifying it. Traditional assertion-based test suites struggle to cover every combination of feature flags, user roles, and edge cases.

To address this, Meticulous records network traffic to mock APIs at replay time, making tests isolated and idempotent. The company modified the browser scheduling engine to remove sources of randomness, such as CPU clock variation and timer interleaving, which reduces test flakiness. An internal indexing algorithm tracks executed lines of code for recorded workflows and selects session subsets that maximize application code coverage. During replay, the system captures screenshots at atomic interaction points to flag single-pixel changes and runtime discrepancies across light and dark modes.

Engineering and Commercial Roles in London and San Francisco

To build out its distributed replay infrastructure, Meticulous is hiring a Software Engineer (£150,000 - £300,000) in London. Joining an onsite team of six, the engineer will construct distributed systems to concurrently replay thousands of sessions in seconds, speed up replay execution while maintaining browser determinism, and derive coverage algorithms that isolate unique edge cases.

The company is also expanding its go-to-market operations across the UK and US. It is recruiting a Founding Account Executive ($280,000 - $400,000) in San Francisco to serve as the first sales hire, partnering directly with engineers on technical deals and traveling on-site to visit enterprise clients. Additionally, the team is hiring a GTM Operator (£60,000 - £120,000) in London or San Francisco to work four days a week onsite in Shoreditch. The GTM Operator will report to the Chief of Staff and CEO, managing growth initiatives across sales, revenue operations, marketing, and customer success.

Role Location Work Mode Salary
Software Engineer London, England, GB Onsite £150,000 - £300,000
Founding Account Executive San Francisco Onsite $280,000 - $400,000
GTM Operator London / San Francisco Onsite £60,000 - £120,000

See all open roles at Meticulous