About us
This group is dedicated to power users who have integrated Claude Code into their core workflow and are driven to stay at the absolute forefront of AI-augmented development. We are moving beyond basic prompts into the realm of high-level intent and sophisticated agentic orchestration.
Our focus areas include:
Advanced Methodologies: Shifting from syntax-heavy coding to intent-based "vibe coding" and rapid prototyping.
The Tech Stack: Deep dives into Claude Code, and the emerging ecosystem of AI agents.
Skills: Tools, Skills, and the Model Context Protocol (MCP) to extend agent capabilities.
Architectural Best Practices: How to structure codebases specifically for LLM readability and long-term maintainability.
Project Showcases: Live demos of what happens when human intuition meets massive AI compute.
Upcoming events
1
Scaling Your Work on Complex Excel Datasets with AI
·OnlineOnlineManage Complex Excel with Agents in Claude Cowork
Live working session designed for operational and business leaders who want to scale how they work on datasets with AI using practical methods that allow you to trust AI agents outputs.
Speaker: Jonathan Yarkoni, CEO @ Latent AI
Join our session for leaders who already use AI and want it to produce work they can actually sign off on. Register to apply.
What You'll Learn
Anyone can get AI to build a spreadsheet. It gets hard when you change one variable three weeks later and need to be certain nothing collapsed, without checking it by hand.
Jonathan Yarkoni, CEO of Latent AI, will walk through how he runs an agent in Cowork and directs it to write its own specifications and unit tests before it builds anything. You leave with a practical way to run agents that build complex Excel spreadsheets, verify themselves, survive changes to their inputs, and do not need line-by-line review every time you touch them.
This is a live teardown of Jonathan's own workflow: how he turned a four-day financial model into four hours of work with agents and output you can sign off on and manage later.## What You'll Do
- Learn the define, refine, test, spec loop and how to run it yourself
- See how to make the agent surface its own gaps instead of you guessing at prompts
- Get the setup that turns one spec into a reusable file you can run again
- Understand where this transfers beyond Excel, to decks, reports, and proposals
- Understand where the method is not worth the effort, so you do not waste time on the wrong work
## FAQs
Is this technical? No. The method runs through Claude Cowork, which embeds agents on a regular desktop and works with your files the way you already do.
Do I need a Claude subscription? Not to attend. To run the method yourself afterward, an active LLM tool is recommended.
I can't attend this session. Can you let me know about future events? Definitely. Register your interest and we'll keep you posted on upcoming sessions.
17 attendees
Past events
6
