Skip to content

Angie Jones

VP of Developer Experience, Agentic AI Foundation · Block

  • 1 episode on Chain of Thought
  • 4 topics covered
  • 1 full transcript
  • Jan 2026 first on the show
Angie Jones

At the time of the conversation, Angie Jones was VP of Engineering for AI Tools and Enablement at Block, the company behind Square and Cash App. On Chain of Thought she explained how her team deployed AI agents to all 12,000 Block employees in about eight weeks using Model Context Protocol and the open-source Goose agent, and the security guardrails that made it safe.

Cover art for How Block Deployed AI Agents to 12,000 Employees in 8 Weeks w/ MCP | Angie Jones

Episodes

What the conversations cover

These notes summarize the episodes featuring Angie Jones, including other speakers in panel conversations. Follow a source link to read the discussion in context.

EP 48 How Block Deployed AI Agents to 12,000 Employees in 8 Weeks w/ MCP | Angie Jones

  • Block deployed its open-source AI agent Goose to all 12,000 employees in roughly eight weeks — with “no playbook,” per VP Angie Jones. Goose started as an internal developer tool built by ML engineer Bradley Axton, then spread organically as finance, marketing, sales, design, and executive-assistant teams asked “why do y’all get to go fast?” Episode takeaways
  • Anthropic showed Block the draft of MCP (Model Context Protocol) before launch — “it was exactly what we needed for Goose.” Block rebuilt Goose as an MCP client and open-sourced it, then fed protocol gaps back from the client side, since most people focus on MCP servers but you also need a client agent to talk to them. Episode takeaways
  • Block refuses to standardize on one model: “no one has won this race yet.” Via a Databricks hosting layer, employees pick from Anthropic’s Claude (favored by engineers for software dev), OpenAI’s GPT (reasoning-heavy roles), and Google’s Gemini. Local open-source models are a niche fan favorite but a “very small percentage,” since agentic use needs a larger model and a machine upgrade. Episode takeaways

In their words

I'm gonna I'm going to take it a step further and do a couple of MCP, because I think that's where MCP really shines, is when you connect multiple systems together to unlock workflows that, like maybe weren't possible before.

Angie Jones How Block Deployed AI Agents to 12,000 Employees in 8 Weeks w/ MCP | Angie Jones · 14:38

Transcript excerpts; ellipses mark omitted words. Some automatically selected quotes have filler words removed.