What is Approxima?

Approxima is a self-serve agentic QA platform that plugs into CI/CD pipelines to autonomously scope, code, validate, and fix software issues end-to-end. Part of YC W2026, the company was founded in 2025 by Ethan Pronev (ex-Jane Street, ex-Citadel Securities) and Ashish Selvaraj (ex-Databricks, ex-Microsoft), both University of Waterloo CS graduates.

When a pull request is opened, Approxima's AI agents open the preview deployment in a real browser, test that the changes actually work (both visually and functionally), identify what is broken, and submit a fix automatically as a follow-up PR. Unlike traditional E2E testing tools that require writing and maintaining test scripts, Approxima requires zero scripting and zero test maintenance — the agents explore and validate applications autonomously.

The platform covers AI bug detection, automated bug fixing, pull request validation, visual regression testing, and agentic QA. Approxima positions itself against competitors like Greptile and QA Wolf, differentiating through its fully automated detect-and-fix loop rather than detection-only approaches.

Key features

Core capabilities this platform advertises.

CI/CD integration
Sandbox validation
End-to-end fix automation
Auto-merge capability
Maintenance automation

Strengths and tradeoffs

What this tool does well, and the limitations to keep in mind.

Pros

Zero-maintenance QA — no test scripts to write or maintain, eliminating the biggest pain point of E2E testing
End-to-end auto-remediation that goes beyond detection to automatically submit fix PRs
Real browser validation against preview deployments catches visual regressions and real-world bugs
Strong founder pedigree from Jane Street, Citadel, Databricks, and Microsoft with deep systems experience
CI/CD native — runs automatically on every PR with no manual triggering

Cons

Plans & pricing

What's included in each plan, and how the tiers compare.

Demo

Contact for pricing

CI/CD integration
Real browser testing
Automated bug detection
Auto-fix PR generation
Visual regression testing

View official pricing page

Common use cases

Engineering teams automating routine dev maintenance

Bug fixing automation
Routine maintenance
CI/CD pipeline integration
Code validation

Using Approxima with Respan

Approxima automates QA testing and bug fixing for AI-powered applications. Respan can monitor the LLM components within those applications while Approxima ensures the overall application quality through automated testing.

Validate AI feature behavior through Approxima testing while monitoring LLM calls via Respan
Catch AI-related visual and functional regressions through Approxima PR testing
Ensure end-to-end quality for AI applications across both code quality and LLM performance