HomeCompaniesDecipher AI

Decipher AI

QA agents that generate tests from session replays and monitor for…

Decipher learns from session replays and writes reliable tests that prevent regressions on all your product flows. Plus get alerted when a bug hits production users.
Active Founders
Michael Rosenfield
Michael Rosenfield
Co-Founder
Michael was most recently a Sr. Product Manager where he led the development of the first native ads format on Google Maps, drove Google Search Ads optimization with $XB revenue growth, and was product lead for Pixel premium tier product line. He started his career at Google as an APM.
Rohan Das
Rohan Das
Founder
Co-founder @ Decipher AI. Previously a staff engineer and manager at Google on the YouTube Homepage, leading the team that rebuilt the Homepage service from scratch to handle billions of requests per day. Passionate about robust yet practical engineering, AI, and elegant design.
Company Launches
Decipher: AI that discovers issues and opportunities in your product
See original launch post

Hey YC! Michael and Rohan here!

TL;DR: Decipher offers real‑time agents that use vision language models to watch every customer sessions to detect, diagnose, and even help ship the fix for anything that blocks user happiness—from silent drop‑offs to full‑blown errors.

Sign Up For Free » getdecipher.com

Book A Call » cal.com/decipher/discovery
Video overview


Why we built Decipher

Shipping speed is up, but visibility into real user experience hasn’t kept pace. Teams still burn days—and sometimes miss problems entirely:

  • Users hitting confusing UI and running into bugs in core workflows
  • Figuring out what exactly people are doing that lead to app errors and how many people are affected
  • Scrubbing thousands of sessions to learn why an important new feature isn’t used

The root cause is signal overload—billions of events, little prioritization. We flipped that on its head: let AI watch the firehose, surface what truly matters to adoption, satisfaction and revenue, and automate the busywork after discovery.


How it works

  • Discover: Real‑time monitoring catches hiccups, drop‑offs, errors—or any custom condition you set, no redeploy needed.
  • Investigate: Ask in plain English and get AI summaries, session clips, and console logs instantly.
  • Act: One‑click Linear ticket or PR draft with all the context so fixes land before metrics suffer.

A feedback loop on every action continuously teaches the model, dialing down noise while accuracy climbs.

Decipher’s basically 1,000 QA interns watching every user session in real‑time and obsessing over improving the user experience. In the first four weeks it surfaced over one hundred issues and UX improvements. Now the team ships features instead of binge‑watching session replays.

— Keith Fearon, Head of Product & Growth, 11x

Issue Explorer

Browse the issues Decipher detects and filter by feature, account, impact, time range, and more. Export straight to Linear.

Session Deep Research

Ask the Decipher agent any question about customers and get detailed responses citing real sessions.

Issue Alerts

Route alerts to your favorite tool, like Slack or email, with AI noise-reduction so teams never miss the signal.

Smart Funnels

Generate funnels based on any user behavior and Decipher will analyze drop-offs to help you understand the root cause.

Where it fits in your stack

  • Setup: Drop‑in SDK or web snippet; optional connectors for PostHog, Sentry, Segment.
  • Workflows: Creates Linear tickets for you. Pings you on Slack and email. And one click copy context to Cursor.
  • Platforms: Web today; iOS & Android are on our roadmap.
  • Security: Data encrypted in transit & at rest and client‑side PII masking.

Integration is < 5 minutes. First insights usually appear within the hour.

Companies of all sizes from small startups to unicorns already rely on Decipher to keep launches smooth and adoption climbing. Their teams ship faster, users stay delighted, and engineering effort goes back to roadmap features instead of reading logs and trying to repro issues.

Get started today, no credit card required at getdecipher.com or schedule a call with our team: cal.com/decipher/discovery

Previous Launches
AI that analyze your session replays with vision LMs to help you understand user behavior, detect bugs and discover insights
Jobs at Decipher AI
San Francisco, CA, US
$160K - $220K
0.50% - 2.50%
1+ years
Decipher AI
Founded:2023
Batch:Winter 2024
Team Size:4
Status:
Active
Location:San Francisco
Primary Partner:Dalton Caldwell