Slack Engineering

What it covers

Slack's blog covers the systems behind a real time messaging product used by many organisations, plus the tools the engineering team builds for itself. Posts from 2026 (checked September 2026) include:

• Shipyard: how Slack built its next generation EC2 platform (July 2026).
• Agentic testing: where agents fit in the end to end testing stack (June 2026).
• Slack AI: the path to multi cloud (May 2026).
• From SSH to REST: a security driven modernisation of Slack's EMR data pipelines (May 2026).
• Managing context in long running agentic applications (April 2026).
• Scalable network probing and HTTP/3 readiness with Prometheus (March 2026).
• How Slack rebuilt notifications (March 2026).

The archive includes widely cited posts on Slack's migration from a large PHP codebase to Hack, its move to a cellular architecture for resilience, real time messaging infrastructure and search.

Why it is worth reading

• Realistic constraints. Slack writes a lot about migrating and scaling systems that cannot be switched off. That is closer to most engineers' reality than stories of building from scratch.
• Clear explanations. Posts tend to explain the problem and the reasoning, not just the final architecture.
• Mix of infrastructure and developer tooling. Testing, deployment and security posts are as useful as the product ones.

Who should read it

• Backend engineers working on real time systems, notifications or messaging.
• Platform and infrastructure engineers, especially those on AWS.
• Engineers adopting AI agents for testing or security who want grounded lessons.

How to read it

• Start with the notifications rebuild post. Notification systems appear in many system design problems, and this is a real one.
• Read the cellular architecture posts in the archive if you care about resilience and blast radius.
• For each post, ask what the old system's failure mode was. The fix only makes sense once you see the problem.

Weak spots

• Posting frequency is roughly one or two posts a month, with gaps.
• Some posts are specific to Slack's internal platforms.
• There is little on frontend or client engineering compared with backend and infrastructure.

Pair it with

• Designing Data-Intensive Applications for the messaging and stream processing theory underneath.
• The Uber and Shopify engineering blogs for more migration stories.

Best for: backend and platform engineers who want honest, practical accounts of evolving a real time product that cannot go offline.