Person profile
CEO & Founder at Shopify.
Tools Tobi Lütke has publicly discussed or demonstrated using. Curated by Magpie's editorial team from podcasts, posts, and public stacks.
Tobi Lütke is the CEO and founder of Shopify, the e-commerce platform that runs millions of merchants. He pushes AI usage as a baseline expectation across Shopify and is a regular voice in the founder community on what AI changes for company building.
Tobi reposts Shopify's official AI Toolkit launch post that names Cursor as one of the supported clients alongside Claude Code, OpenAI Codex and VS Code. Shopify provisions Cursor across the engineering org. Bessemer's write-up of Shopify's engineering playbook records that the company ordered an initial 1,500 Cursor licenses and quickly bought another 1,500. Cursor sits in his stack as the default IDE for AI-assisted coding.
Tobi posted a thread in January 2026 about running Claude on the USB stick from his annual MRI scan and asking it to build him an HTML-based viewer because the bundled software was Windows-only. He shipped the viewer in roughly one prompt, then added scrollable scan stacks, zoom and spine annotation in follow-ups. The post crossed 7.5M views and became one of the most-cited consumer examples of building one-off software with Claude.
My annual MRI scan gives me a USB stick with the data, but you need this commercial windows software to open it. Ran Claude on the stick and asked it to make me a html based viewer tool. This looks... way better.
Claude Code is named in the Shopify AI Toolkit launch that Tobi reposted, listed first among the supported clients. The toolkit is an MCP server that exposes Shopify's Admin API and GraphQL schema, so any Claude Code session can manage a store from the terminal. Tobi positions Claude Code as the agent surface for store-level work, sitting alongside Cursor as the IDE.
OpenAI Codex is listed in the Shopify AI Toolkit launch post that Tobi reposted. Codex shows up alongside Claude Code and Cursor as a supported MCP client, slotting into the per-developer-choice approach the engineering org takes. Tobi has been explicit that Shopify lets engineers pick the agent that fits their work rather than mandating one client.
the Shopify AI Toolkit is here manage your store with your favorite agent Claude Code, Codex, Cursor, VS Code, and more
GitHub Copilot is part of the official set of AI coding tools Shopify provides every engineer, alongside Cursor and Claude Code. Bessemer's coverage of Shopify's engineering playbook names Copilot in the mix Tobi sponsors at the company level. He hasn't posted about Copilot personally the way he posts about Claude or Cursor, so this is a stack-level rather than personal-driver signal.
Shopify engineers use a wide mix of AI coding tools, including Cursor, Claude Code, GitHub Copilot, OpenAI Codex, and experimental tools from Gemini.
Tobi spent a 32-minute video interview in November 2025 demonstrating his Raycast setup with custom hotkeys, window management and scripts. He went further: he personally funded a $5,000 bounty for a developer to land a Chrome --focus CLI flag so a single command could open or refocus a web app, specifically to compose with Raycast. He frames Raycast as the keyboard-first surface that ties his Mac workflow together.
I just happily sent the $5k bounty to Hjanuschka for landing a new feature in Chrome that was born in the linked discussion thread: Chrome now gets --focus cli parameter which allows you to open a website / webapp from the command line (or Raycast) OR focus it in an existing
MCP is the protocol Tobi reaches for when he wants his own tools to compose with Claude. He built QMD, his personal local semantic-search tool, to expose itself as an MCP server so Claude Desktop and other clients can query his notes directly. Shopify's own AI Toolkit ships as an MCP server too, with Tobi reposting its launch. He treats MCP as the wiring layer for stitching his personal and company stacks together.
QMD, Tobi's personal local-search tool, runs three Qwen models under the hood: Qwen3-Embedding-0.6B for embeddings, qwen3-reranker-0.6b for LLM reranking, and a fine-tuned Qwen 1.7B for query expansion. The whole stack runs locally via node-llama-cpp with GGUF model files. It's a useful signal that Tobi reaches for open-weights models when he wants full local control and no cloud dependency.
Last researched May 12, 2026
Editorial note
These tools have been mentioned by the people featured in podcasts, interviews, posts, and other public media. Source links and citations are included wherever possible. Magpie AI has no direct relationship with the people on these pages. Spot a missing tool or a wrong attribution? Suggest a correction →
Photo: X