Recording of my London Gophers talk is now available:
www.youtube.com/watch?v=--M5...

Had a great time at LondonJS tonight - especially enjoyed the @flagsmith.bsky.social journey of bootstrapping their open source project to scale.

Lesson: no gimmicks…but fireworks when required! 🎇

At every meetup or after a talk someone asks: "why not use MCP, why orra?"

MCP = connect LLMs to systems
Orra = coordinate agents in production

Need state coordination, failure handling, audit trails? That's a lot to rebuild.

They work better together!

Comparison: github.com/orra-dev/orr...

Our latest technical deep dive by @ezodude.bsky.social is getting some love over at @dev.to

Give it some more 💜: dev.to/ezodude/self...

🚀 Thread: After months in stealth, thrilled to finally unveil what @ezodude.bsky.social and I have been crafting!

Introducing Orra - our open-source glue layer for multi-agent applications. We're coming out of stealth today! 🧵

🚀 Thread: After months in stealth, thrilled to finally unveil what @new_nomad and I have been crafting!

Introducing Orra - our open-source glue layer for multi-agent applications. We're coming out of stealth today! 🧵

Exciting times! Running one of our examples through DeepSeek, we can see the reasoning = new possibilities! Geeking out with @ezodude.bsky.social

Excited to see Pydantic's new pydantic-ai project!

Some nifty features like streamed responses, agents as dependencies, and vanilla Python control flow.

Can't wait to see what the Python community builds. When teams are ready to move their agents to production, we'll be here to help them hum! 🚀

Just shipped our Python SDK! 🐍

Started with JS to serve the 13.8M devs often overlooked in AI tooling. Now, we're bridging both worlds.

Found that building the Python SDK improved our JS offering too—cross-pollination for the win!

#AIAgents #StealthStartup

👋 Hi, I’m Ezo! I’m new here, so here’s a quick intro:

Experienced developer wrangling LLMs and AI agents. Co-founder of a startup in the agentic applications/dev tool space and run an AI consultancy.

Hip-hop head with a growing record collection, and I drink coffee like it’s fine wine! ☕🎶

What’s happening this week? Here’s what I have on my laptop, in the hopper, and in my ears 👇

Coding 👨‍💻: Node, Python, testing with AsyncAPI Specification
Coffee ☕: Banko Gotiti, Ethiopia / Washed from Artisan Roast
Beats 🎧: www.youtube.com/watch?v=NQFR...

This is fun, which city can get you to Ramen Profitability asap? 💡The Ramen Index can help: www.producthunt.com/posts/the-ra...

1/7
New LangChain survey drops truth bombs about AI agents in business! 1,300+ professionals reveal it's not just hype - companies are actually using these in production. A whopping 51% already deployed, and 78% are gearing up to launch soon! 🚀

2 weeks ago, I held white-boarding sessions in London, collaborating on LLM & AI agent architectures—great discussions were had! 🤓

This week, I’m back:
📍 Nov 19-20: Old Street
📍 Nov 21: Holborn

DM for details - let’s brainstorm or tackle challenges! 🚀

#AI #LLM #TechLondon

At the MLOps AI Agents conference, I learned GPT costs dropped 98% to $0.24 per 1M tokens! Incredible!

But agent costs are up 23% ($0.26/question). Why? Agents think like consultants: more questions, higher costs. 😄

How are you balancing efficiency & intelligence with your AI agents?