# How YC Hit 90 % Coverage on Its Mission‑Critical Applications Portal

A conversation with

Sneha Sivakumar

CEO of Spur

- **Company**  
  Y Combinator is a startup accelerator that provides seed funding, advice, and connections to help early-stage companies grow.  
- **Industry**  
  Venture Capital / Accelerator  
- **Company Size**  
  51-200  
- **Founded**  
  2005

## Challenge
For an accelerator that reviews thousands of startup applications on tight deadlines, even a tiny glitch can block founders and erode YC's brand. The applications team faced major QA hurdles: branching multi-step logic with hundreds of unique applicant paths, lean engineering bandwidth with just two engineers, and surge-deadline volume with tens of thousands of founders submitting within the final 24 hours.

## Solution
Spur's AI-powered agents simulate real applicants end-to-end, achieving 90% coverage in two weeks. The team uses a record-to-test workflow where they record a flow and Spur writes the test. Adaptive applicant agents handle every field and state change with zero script maintenance, and instant Slack alerts ensure issues are fixed before founders notice.

## Results
90% Application-Flow Test Coverage  
- **90%** Application-Flow Test Coverage  
- **100+** Manual QA Hours Saved per Batch

> "Spur is shockingly easy to use—no coding, just plain English. Simply describe what you want to test, and Spur handles the rest. Onboarding was a breeze, and Anushka and Sneha are fantastic to work with."

Eve Bouffard  
Product, Y Combinator

## High-Volume Deadlines, Lean Engineering Team
For an accelerator that reviews thousands of startup applications on tight deadlines, even a tiny glitch can block founders or flood support and erode YC's brand. The applications team faced three major QA hurdles:

- **Branching, Multi-Step Logic:** Dozens of conditional sections generate hundreds of unique applicant paths - far too many for manual test scripts.

- **Lean Engineering Bandwidth:** With just two engineers maintaining the portal and four batches a year, the team couldn't spare days for bug-bashes before every deadline.

- **Surge-Deadline Volume:** Tens of thousands of founders submit within the final 24 hours, so performance and reliability must hold up when traffic peaks.

## Spur launches thousands of Agents each batch to emulate founders going through the flow
Spur's AI-powered agents simulate real applicants end-to-end:
- **Rapid, comprehensive testing** - 90% coverage in two weeks  
- **Record-to-test workflow** - Ev records a flow, Spur writes the test  
- Adaptive applicant agents that handle every field/state change with zero script maintenance  
- Instant Slack alerts so issues are fixed before founders ever notice

Critical bug caught the Sunday night before peak submissions, preventing an outage for thousands of founders.

## Ready to transform your testing?
Schedule a demo to see how Spur can handle all your QA, save development time and prevent costly bugs.