The Workshop.
Notes on browser automation, local AI and building in public.
Latest articles
Showing Local AI
Local AIRunning LLMs inside a Chrome extension: WebLLM, transformers.js and the MV3 traps
What it took to run AI models inside a Manifest V3 extension: three runtimes, two Web Workers, a WebGPU probe with a WASM fallback, and the CSP rule that silently blocks your runtime.
Local AITool calling for small on-device LLMs that don't support tool calling
Models running in the browser give you text in and text out, so agent frameworks refuse them. How we added prompt-based tool calling with a model-agnostic parser: one format asked for, many accepted, and three rules that keep it safe.
Local AIWe rebuilt RAG in a browser tab. Our first bug: every vector was empty.
Our knowledge bases embed, store and search your documents inside the browser. Before rebuilding them, an audit found that no knowledge base had ever had a single searchable chunk. Ten RAG lessons from that rebuild, written so you can check your own pipeline.
Local AIWhy vector search misses exact terms, and what hybrid search fixed (measured)
Ask a vector search for ERR_4521 and you may get ERR_4512. How we added BM25 next to vectors with reciprocal-rank fusion in an in-browser knowledge base, what we measured, a SurrealDB full-text gotcha, and a filtering bug that gave small collections zero hits.
Local AIAWFlow wants to be the OpenRouter for AI that runs in your browser
Three incompatible browser AI engines, a different model format each. AWFlow sits in front of all three so you choose a job, not a library. Where we are and why the web will get there.
Local AIDo you need a big AI model? A one-hour test that tells you
The biggest model is a habit, not a requirement. How to tell which of your automation jobs a small, free, on-device model can handle, with a simple test on your own inputs that takes about an hour.
Local AIFree AI vs API-key AI: what an automation really costs
How to estimate an AI step's cost before you build it, why the page you feed in is usually the bill, and a simple design that sends only the hard work to a paid model. With a calculator for your own numbers.
Local AIThe airplane-mode test: the only privacy proof for AI you can run yourself
Every AI tool says it respects your privacy. One test settles it: switch off the internet and see if it still works. What offline AI really requires, how a workflow can fail the test, and how to run it in five minutes.
Local AIWe need an OpenRouter for the AI that runs in your browser
Cloud AI became easy once routers made models swappable. AI that runs inside your browser is still split into islands that cannot share models. Why that is, what a router for on-device models must do, and what we are building.
Local AIWebAssembly, explained: how a browser tab became a small computer
A browser tab can now run a real database, Python and AI models, with no server. The reason is WebAssembly. What it is in plain words, what it is good and bad at, and how you can spot it yourself.
Local AIWebGPU vs WebGL vs WebAssembly: the day the fast path gave us wrong answers
There are three ways to run an AI model in a browser. What each one is, how libraries choose between them, and a measurement of ours where the faster option silently returned useless results.
Local AIRun AI Nodes Directly in Chrome with WebLLM and Chrome AI API
Ollama requires a server. Cloud models require an API key. WebLLM and Chrome’s built-in AI API require neither — they run LLM inference entirely inside your browser tab. Here’s what that unlocks for no-dependency AI workflows.
Local AIBest Free LLMs for Workflow Automation: Mistral, LLaMA, Gemma
You don’t need a paid API to build AI-powered workflows. Mistral, LLaMA, and Gemma now rival GPT-3.5 for the most common automation tasks. Here’s how to choose the right free model for summarization, classification, and extraction.
Local AIHow to Use Ollama in a Workflow Automation Tool (No Cloud Needed)
Every AI workflow tutorial says ‘add your OpenAI key here.’ Ollama lets you skip that entirely. Connect a local LLM to your workflow automation tool and run AI steps on your own machine — private, offline, and free.
Local AIStop the Tab-Switching: Build Your Own AI Webpage Summarizer
Copying text into a chatbot tab breaks your focus. Build a four-node workflow that summarises any web page right where you read it, with a model that can run on your own device. Video tutorial included.
No article matches that search.
One letter a month. No noise.
New articles, templates worth stealing and what shipped. Unsubscribe in one click.