At Replay.io, we’re building the first deterministic browser. If you can solve this one nearly impossible problem, you unlock Runtime Replay.

Here are five things that are only possible with Runtime Replay

  1. Perfect reproducibility Say goodbye to “works for me” because once you’ve captured a bug, you can debug it later as if it’s failing consistently on your laptop.
  2. Collaborative devtools Once you’re able to replay the runtime in the cloud as if it’s running on the original computer at the original point in time, you make it possible for anyone to investigate the issue as if they were there when it happened.
  3. Time travel enabled devtools Traditional browser devtools are limited to inspecting what’s available in real time. When the runtime is able to efficiently replay, it’s possible to design a debugging environment that more closely resembles a visualization tool.
  4. Dynamic analysis at scale There are two branches of software analysis: static and dynamic. For the past 50 years we’ve primarily focused on static because it’s the only one that we can scale. Replayability makes it possible to perform general purpose control flow and data flow analysis such as detecting new limiting paths.
  5. Replay environments With the ability to replay comes the ability to simulate counterfactuals such as a faster network calls, modified code, different user interactions that automate the debugging journey and down the line power the reinforcement learning environments for state-of-the art model training.

Where are we now?

We are four years in, but in many ways we’re just getting started.

As a company, we’ve primarily focused on our initial cohort of customers and have users who live in and could not live without Replay DevTools. We’ve helped teams like Glide, TableCheck, and NextJS transform their bug reporting process. And we’ve helped teams like Weights and Biases, Metabase, and <redacted> shake up their test suites and achieve previously unbelievable test run success rates.