Recording: The Future of Design-to-Code Workflows with TJ Pitre
Missed our session with TJ Pitre, creator of Figma Console MCP? The recording is now available for Lead by Design members. This was one of the most practical sessions we’ve had yet on where AI, design systems, and design-to-code workflows are heading. A few standout insights from the conversation: - AI is only as good as the context it receives. TJ made the point that the real value isn’t just in components, tokens, or data structures — it’s in the meaning, rules, intent, and context flowing through the system. - Design systems need to be AI-ready before AI can really help. Otherwise, AI fills in the gaps, makes judgment calls, and introduces drift. - Designers may increasingly contribute the first coded draft. One of the most interesting ideas was that designers can now move beyond handoff and help create the first meaningful version of a component in code, while a context engineer helps steward quality, standards, trust, and integration. - Figma Console MCP is not just about generating code. TJ showed how it can support diagnostics, reporting, design generation, coded component generation, and design-code parity checks. - The future is not “AI replaces the team.” It’s more likely a model where humans stay upstream, define the constraints, and orchestrate the workflow, while AI executes inside those boundaries. The live demo was awesome. TJ showed: - a workflow for generating a coded component from an existing Figma component - a blank-page design generation workflow using the existing component library - how parity checks can help keep design and code aligned - how richer annotations and documentation improve downstream AI output A very big thank you to TJ for such a thoughtful and practical session. Once you’ve watched it, I’d love to hear:What part of this workflow feels most immediately useful in your own team or practice?