Lessons learned from two years of building AI agents at PostHog

When to reach for a well-known URI — and the traps that catch most protocol designers.

sapiens

The reasoning layer that sits above the database, and that increasingly treats the database as infrastructure, is where a new generation of companies is being built.

To build useful and usable AI-powered systems, our understanding of users’ needs and our design judgement must be encoded into well-defined evaluation criteria.

Learn what prompt engineering is and how to create prompts so AI is best able to help you.

Notes from my Thoughtworks colleagues on AI-assisted software delivery

Your AI coding assistant can see your code but not your organization. The Reach pattern is a personal CLI that borrows your real browser sessions to call web APIs on your behalf.

Charles-Axel Dein's personal website.

Everyone wants to be successful.

The big news this morning: Astral to join OpenAI (on the Astral blog) and OpenAI to acquire Astral (the OpenAI announcement). Astral are the company behind uv, ruff, and ty—three …

tl;dr: When you’re matching capabilities to applications — AI tools to roles, research teams to industry sectors, platforms to use cases — the

Learn how to proxy fetch() requests in Node.js, Deno, Bun, and Cloudflare Workers to better monitor and control your server-side traffic.

If your team members dread the notification that they’ve been added as a reviewer on a pull request, I think the following guidelines can help.

Why every role seems like an engineering role now (and what it means for you)

From DeepSeek-V3 to Kimi K2: A Look At Modern LLM Architecture Design

A multitudes study which followed 500 developers found an interesting soundbyte: “Engineers merged 27% more PRs with AI - but did 20% more out-of-hours commits”.

Learn how to build flexible, mathematical typographic scales using :heading(), sibling-index(), and pow() for cleaner CSS design systems.

The real test isn’t whether your component works on your current page. It’s whether it works when someone else uses it.

Job postings for Forward Deployed Engineers (FDEs) have surged, but many professionals don’t want the role because it’s more like solutions engineering than software development.

I’ve been thinking about scale the same way I’ve been thinking about speed: as something Silicon Valley treats like a moral good when it’s actually just a strategic choice. The in…

I was reading a post on Anthropic’s blog about AI product management, and the author talked about speed as if it’s a virtue. She never said why speed was supposed to be a good thing. Sh…

You, me, and everyone we know could get better at the art of giving feedback. The debate about feedback at work - how to do it, how not to do it - is ongoing. The jury may still be out, but there’s...

(I used to do more of these visual build posts.

Paul Rand, org designer. I also especially like how the essay's title hints at *willfulness* and the last line points at an economic outcome. #nice

Real strategy means choosing between two good options and accepting all the consequences--even the painful ones you don't like.

How to choose between modals and pages, when to avoid modals, and how to determine the right level of interruption or navigation. Brought to you by Smart Interface Design Patterns, a **friendly video...

We’ve all heard of those network effect laws: the value of a network goes up with the square of the number of members. Or the cost of commun...

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

Companies are ditching junior engineers en masse. They'll tell you it's because AI writes better code. They’re right...but also missing the point.

UX faced instability from layoffs, hiring freezes, and AI hype; now, the field is stabilizing, but differentiation and business impact are vital.

In Part 4 of The Future of Design Series we reveal three paths to sovereignty for designers in the AI era: Brand Sovereignty (scaling creative work independently), Product Sovereignty (building AI-native...

Claude Code now supports agent teams - coordinated swarms of AI agents that research, debug, and build in parallel. What was feature-flagged is now real. H...

Learn how to write effective specifications for AI coding agents to improve clarity, focus, and productivity in your AI-driven development workflows.

Run AI coding tools in autonomous loops with Ralph Wiggum. 11 tips for AFK coding: scope, progress tracking, feedback loops, and shipping code while you sleep.

After reviewing dozens (hundreds?) of operating systems in depth, I’m becoming increasingly convinced that the thoughtful design and deployment of constraints is a superpower.

A blog about software craft and quality

More lessons learned from 14 years of engineering at Google, focusing on what truly matters beyond just writing great code.

Learn to make HTTP requests in Node.js using built-in fetch(), http/https modules. Covers POST, authentication, streaming, and testing with code examples.

What you need to know about how to track and communicate the value of your design system

Optimizing LCP (Largest Contentful Paint) is not just reducing the file size of your images, there's a lot more that goes into optimizing the loading strategy - and you can only grasp where the bottleneck...

The barrier to entry for building software has collapsed. The barrier to building something that matters hasn’t moved an inch.

A guide on how browsers work.

The Open Web means several things to different people, depending on context, but recently discussions have focused on the Web's Openness in terms of access to information -- how easy it is to publish...

Who Should Own the Truth?

If you're not caught up on the current hypewave, coding agents like Claude Code and Cursor have latched onto skills recently. Like with all of these waves, everyone has immediately proclaimed they solve...

Who, what, where, when, why, how

Quite quickly, AI products have transitioned from models behind the scenes powering features to people talking directly to models (chat) to models deciding whic...

Why do federated design systems keep failing? I watched it happen twice at Spotify. Here's what actually happened, where they broke down, and why centralised models usually work better.