AI agents can generate code faster than traditional CI can validate it. We need to change how we think about CI to keep delivery fast and maintain trust.

When boxes are MIA, the multi-armed bandit saves the day.

Perspectives from engineers on software engineering, AI, infrastructure, security, and the problems we’re working through.

Depot Metal's autoscaler grew from a simple pressure loop into a topology-aware system for compute and storage. Next, it must predict demand and converge on desired capacity.

We built a production-shaped load test to find control-plane limits before moving more GitHub Actions traffic to Depot Metal. Each run made our scaling changes measurable.

Join Depot and Inference for a casual SF Tech Week co-working day that turns into a happy hour with demos and networking at the Inference office by the…

YouTube video by Depot

Depot CI runs GitHub Actions workflows without GitHub’s control plane, keeping CI online during outages and making workflows portable to new forges like Origin.

Taking the Sandbox TTI benchmarks one step further for fun.

Our microVMs launch in under a second, and assembling their block devices in parallel put LVM's roughly 100ms volume group-wide lock directly in the critical path. We kept the kernel's device-mapper, ...

Agents can investigate, plan, and implement faster than I can, but reading a final diff is not the same as understanding why the system works the way it does. I own how the system behaves in productio...

Most of a large container's startup goes to pulling and decompressing files it never reads to boot. SOCI makes the image seekable so the runtime loads only the bytes it needs, and Depot writes that in...

GitHub isn't just falling short on performance and reliability. The whole collaboration paradigm is the wrong shape for how we build software now.

Taking the Sandbox TTI benchmarks one step further for fun.

Formal verification never seemed worth it: writing a faithful TLA+ spec took too long. But now agents can write the spec from your code. We tried it on the rebuilt Depot Registry garbage collector, an...

Keeping our AI assistant working well is a human responsibility at Depot. PostHog's AI engineering features turned our never-ending reviews into 30 minutes a week.

A team process without a principle underneath it is a real problem. Here's why the "why" is worth protecting as you scale.

Launching today in early beta, Origin code hosting features a Depot app to run your existing GitHub Actions workflows on Depot CI for better performance and reliability without GitHub.

Our microVMs launch in under a second, and assembling their block devices in parallel put LVM's roughly 100ms volume group-wide lock directly in the critical path. We kept the kernel's device-mapper, ...

To make recurring test failures easier to reason about we needed to include historical context. Here's how we derive a stable identity for each test and a fingerprint for each failure, so you can tell...

Agents can investigate, plan, and implement faster than I can, but reading a final diff is not the same as understanding why the system works the way it does. I own how the system behaves in productio...

AI can produce code faster than you can review it, but you're still the one on call when it breaks. You need an AI-assisted workflow you trust: linter guardrails, reference tests the agent can copy, t...

GitHub isn't just falling short on performance and reliability. The whole collaboration paradigm is the wrong shape for how we build software now.

Report JUnit from your Depot CI or Depot GitHub Actions jobs to see failures in context. A single command runs your suite and uploads the report. Split suites by historical timing and they'll automati...

Most of a large container's startup goes to pulling and decompressing files it never reads to boot. SOCI makes the image seekable so the runtime loads only the bytes it needs, and Depot writes that in...

Formal verification never seemed worth it: writing a faithful TLA+ spec took too long. But now agents can write the spec from your code. We tried it on the rebuilt Depot Registry garbage collector, an...

Keeping our AI assistant working well is a human responsibility at Depot. PostHog's AI engineering features turned our never-ending reviews into 30 minutes a week.

GitHub isn't just falling short on performance and reliability. The whole collaboration paradigm is the wrong shape for how we build software now.

A team process without a principle underneath it is a real problem. Here's why the "why" is worth protecting as you scale.

Split a test suite across matrix jobs with historical JUnit timing data, then run and report every shard.

To make recurring test failures easier to reason about we needed to include historical context. Here's how we derive a stable identity for each test and a fingerprint for each failure, so you can tell...

AI can produce code faster than you can review it, but you're still the one on call when it breaks. You need an AI-assisted workflow you trust: linter guardrails, reference tests the agent can copy, t...

Formal verification never seemed worth it: writing a faithful TLA+ spec took too long. But now agents can write the spec from your code. We tried it on the rebuilt Depot Registry garbage collector, an...

Depot is joining the Stripe Projects developer preview. Developers and their agents can now set up Depot from the terminal: create or link an organization, pick a plan, and get a token back that unloc...

How PlanetScale scrapped Buildkite and runs CI in half the time with Depot

Keeping our AI assistant working well is a human responsibility at Depot. PostHog's AI engineering features turned our never-ending reviews into 30 minutes a week.

A team process without a principle underneath it is a real problem. Here's why the "why" is worth protecting as you scale.

Source control has become the bottleneck at the top of the delivery loop. So we built our own: a diskless git server backed by blob storage that scales horizontally, mirrors your GitHub repos, and plu...

Depot Metal is our next-generation compute and storage platform. Depot CI and Sandboxes are now running on it, with benchmarks showing workloads 30% faster and more optimizations on the way. You don't...

To make recurring test failures easier to reason about we needed to include historical context. Here's how we derive a stable identity for each test and a fingerprint for each failure, so you can tell...

How PlanetScale scrapped Buildkite and runs CI in half the time with Depot