Launching today
LiveReview
Free, Risk-Aware Code Review: Auto Approve Low-Risk Changes
183 followers
Free, Risk-Aware Code Review: Auto Approve Low-Risk Changes
183 followers
Most AI review tools flag issues. LiveReview promises production results: fewer incidents, fewer rollbacks, less senior reviewer load. Every PR is risk-scored. The lowest-risk 30-50% are safely auto-approved; the risky ones go straight to the engineers who should see them. The approach is inspired by Meta's highly successful internal AI review system, where production incidents fell 50x and rollbacks 3x. LLM + incident-trained ML + code-graph analysis. Free and self-hostable.
Free Options
Launch Team
Render Sandboxes: a computer for every agentGive every agent its own machine. Starts in under a second
Promoted
Very nice, congrats on the launch!
@nathan_tarbert Thank you for the support :)
@nathan_tarbert Thanks Nathan <3
how does an incident get linked back to a PR? ours show up days after the deploy, in Sentry or in a saturated db pool, and rarely point at one commit. do you read Sentry/PagerDuty or do we tag reverts by hand?
@raphael_millet2 We have both UI and API webhooks for marking production incidents/reverts. So both are possible, depending on your exact setup (happy to work with you and configure it). We do want to develop sentry/pagerduty native connections as well (on the way).
auto-approving the low-risk PRs is the useful part. when it routes a riskier one, does the engineer see the risk score breakdown or just a flag?
@neerajnathany Detailed risk score is available; almost every piece of the final result can be inspected. This is useful for tweaking the system over time.
cool launch, congrats!! if a one-line config change hits the payments path, does LiveReview flag it from the code graph or could it get auto approved as low risk?
@igorblink thanks for engaging. For things such as auth, security and payments -- it'll catch things as high risk and flag it for humans for review.
Each organization can also define it's risk appetite as a percentile, so a regulated enterprise can have a conservative setting whereas a startup can take more aggressive risks and our tool helps each org set the risk threshold they wish to take on.
There are also individual sliders for various risk dimensions in correctness, compliance, security, performance, scaling, cloud costs, etc.
@shrsv amazing :)
Auto approving the lowest risk third and sending everything else to the right engineer is the version of this that would survive a real Monday, since another tool that flags fifty things a day just trains us to ignore it.
@karimbenkeroum Perfectly captured, Kareem. It's due to the increase in number of PR and also size of each PR, attention is at a premium. Attention must be guarded and directed at the most important parts that deserve it.