Hey guys, I just thought about AI and Agents within small companies (less than 10 employees). As Claude etc. has some serious problems with the European GDPR, I was wondering if anyone is using Microsoft Copilot and Studio and could tell me his opinion about it in comparison to Claude? Like, I am not talking about software dev, but general AI agentic work. Like, Getting shit done. Do you also using the folder structure in Copilot? Is it worth a try? Can you connect external MCPs? Cheers to everyone!

This might be a basic question, but I'm trying to figure out the best way to structure my marketing folders when juggling multiple products. Would you recommend a Task-Centric or a Product-Centric approach? For context: I work on very different projects (course, Consulting, and now also small Software/Apps), but they all share the exact same marketing processes, like social media marketing. This initially made me lean toward a Task-Centric structure. However, the downside is that all the specific assets and data regarding a single product end up scattered everywhere. On the other hand, I realized this problem is actually much more far-reaching. I have a single, overarching website that hosts multiple products. Because of this, many assets (like website copy, global funnels, or general branding) have no direct reference to just one specific product. In a strict Product-Centric structure, there's no logical place to store these shared assets. Furthermore, the massive benefit of the Task-Centric approach is having a single, centralized place to store instructions (SOPs) and learnings, so that those skills continuously improve across all projects in the company. Because of this dilemma, I came up with a third option: a Hybrid Structure. In this setup, I store all overarching Learnings, Skills, and Templates (including global website/brand assets) on the very first layer (the root folder), while keeping the actual day-to-day marketing and ad assets grouped by Product. (Essentially mixing Example 1: Content Creator and Example 2: Freelancer from Lesson 3.2). What are your opinions on how to structure this best? What would you improve? Product-Centric: Marketing/ ├── Product 1/ │ ├── Scripting & Angles/ │ ├── Video Production/ │ │ ├── 01_Projects/ (Premiere, AE) │ │ ├── 03_Footage/ (Raw video) │ │ └── 04_Exports/ (Final MP4s) │ ├── Carousel Production/ │ └── Learnings/ └── Product 2/ ├── Scripting & Angles/ └── ... Task-centric: Marketing/ ├── Scripting & Angles/

Hey I'm working in Antigravity and using Claude Code and hitting my limits daily. How do you handle hitting the limit? Do you have a workaround to deal with it? Certain tasks handled by other LLM? Complete switch after the limit is hit? What LLM do you switch to? I'm a bit nervous to use another LLM... afraid of the quality of the work it will do. A lot of questions in one post. 🤯 Please share your thoughts.

Most property analysts still manually dig through Grundbuchauszüge and Exposés. It's slow, error-prone, and doesn't scale. So I built Kimora — a fully local LLM-powered document analysis tool designed for German real estate workflows. What it does: ✅ Analyzes German property documents (Grundbuch, Energieausweis,Exposés..) ✅ Extracts structured data automatically using a local AI model ✅ Detects missing or outdated documents and flags risks automatically ✅ Tracks all activities, tasks & work done per property object ✅ Keeps a full history log so nothing falls through the cracks ✅ Generates clean PDF reports — no copy-pasting ✅ Prepares and publishes Exposés directly to social media (Instagram, X, Linkedin..) ✅ 100% local & GDPR-friendly by design ✅ Fully Dockerized — spin it up anywhere in minutes ( I love Docker 🐳 )

Every time a new model or agent framework drops, I feel the urge to burn everything down and rebuild. New tool, new “agent,” same problem: my workflows, prompts, and SOPs end up glued to the tool instead of the work. What finally started working was flipping the mental model: The folder is the app. The AI is just the worker that knows how to navigate it. Under the hood I’m using a simple architecture that does two things: Separates what should be code, what should be rules, and what actually needs AI Structures everything so any model can plug in later without a rebuild I think of it as 60 / 30 / 10 on top of a 3‑layer workspace. Step 1: 60 / 30 / 10 – decide where AI actually belongs For every task, I force it into one of three buckets: 60% – Automations (no AI at runtime) Same input → same output every time. Scripts, API calls, cron jobs, database writes. AI can help write the script, but the script runs it. 30% – Rule engines (templates / decision trees) Clear if/then logic, multiple branches, but all map‑able. Triage, routing, form logic, templated replies. Once the tree exists, it runs without “thinking.” 10% – Prompts (actual AI work) Summaries, syntheses, ambiguous inputs, creative drafting, nuanced judgment. This is the only layer where I want the model to “think.” If I can’t decide, I decompose the task until each step is clearly one of those three. That alone kills a ton of “let’s just build an agent” impulses. Step 2: 3‑layer folder architecture Once tasks are classified, everything lives in a single workspace: Layer 1 – The map (router) One root markdown file, e.g. _router.md or WORKSPACE.md: Explains the business / project in plain language Lists folders and what they’re for Defines naming conventions (this is your “schema”) Has a routing table: “If task = X, read these files, skip those, write output here, load these skills.” When I start a session I don’t paste giant prompts. I say: “Read _router.md and follow it.” Layer 2 – Task workspaces by layer