The Birth of OTTO: How We Built an AI Crew for Our AWS Community
Every user group starts small.
A few passionate people.
A few meetups.
A few pizzas.
But if you’re lucky — if you care deeply, show up often, and invite others to lead too — it grows into something much bigger.
Today, our AWS User Group Vienna in Austria has over 2000 members.
We have five incredible organizers and User Groups in Vienna, Linz, and just founded an additional Women’s User Group in Vienna.
And the movement is still growing.
But before all of that — before the scale and the team —
there were just two of us: Philipp and me, but not always at the same time.
Philipp had led the user group alone for a while, and when I took over, I carried the responsibility solo for several years.
Eventually, Philipp rejoined, and together we began sharing the weight of the growing community.
Behind the scenes, an endless mountain of invisible work awaited us both.
🎮 Why the Slides Look Like The Sims
We wanted our presentation to feel relatable and fun, so we styled the slides like The Sims. Just like building a house in the game, organizing an AWS User Group meetup involves assembling pieces, managing resources, and dealing with repetitive tasks. The Sims metaphor perfectly captures the mix of creativity and routine that goes into creating something meaningful — and how automation, like OTTO, can make it all a little less tedious.
🧠 From “We Got This” to “We Need Help”
Both of us had been there:
Running meetups alone.
Chasing speakers.
Begging venues.
Writing emails at midnight.
Publishing events while juggling full-time jobs and side projects.
At first, it felt manageable.
But growth — beautiful as it is — amplifies friction.
And friction, over time, wears you down.
Even when Philipp and I teamed up officially, it became clear:
Two people can’t scale a community past a certain point alone.
We didn’t want to slow down.
We didn’t want to lower our standards.
We didn’t want to burn out.
We needed backup.
🤖 Not More Humans. A Crew.
We didn’t dream of another project management board.
Or a calendar full of Zoom calls.
Or spreadsheets nobody updated.
We dreamed of something smarter.
Something alive.
A crew — but made of AI agents.
Each agent would catch a different ball:
- A Speaker Finder to surface great speakers automatically
- An Agenda Producer to draft meetups
- A Q&A Agent to answer the 100th “what is the AWS User Group?” email
- A GitHub Updater to track our tasks
- A Header Generator to create visuals without fiddling with Canva
Not a monolithic app.
Not another to-do list.
A modular, intelligent crew.
🔥 Building OTTO
We sat down and started sketching it out —
sometimes in late-night sessions, sometimes between other projects, sometimes just dreaming aloud.
It felt crazy.
It felt necessary.
It felt right.
We chose:
- CrewAI to organize the agents
- Amazon Bedrock to host custom-tuned models
- InstructLab to train the agents on our domain knowledge
- Slack as our mission control center
We didn’t want a better tool.
We wanted a better way to work.
🌱 Where We Are Now
Today, we’re not alone anymore.
The AWS User Group Austria has a full team of passionate leaders.
We’ve expanded meetups beyond Vienna into Linz and other cities.
And soon — very soon — we’ll deploy OTTO into our real environment.
Every agent we built will be tested live:
- Real event planning
- Real speaker scouting
- Real GitHub syncs
- Real Slack conversations
We’ll gather feedback.
We’ll tune and retrain the agents.
We’ll make OTTO better with every real-world cycle.
Because that’s the secret:
Communities aren’t static. Neither is OTTO.
🚀 What’s Next
This is just the first part of OTTO’s story.
Next, we’ll dive into how we designed the architecture:
the way the agents interact, how we trained knowledge into them,
and why starting small and modular made all the difference.
➡️ Stay tuned for Part 2: “Turning Chaos into Crew: Designing OTTO’s Architecture”
📬 Curious? Inspired? Want to build your own Crew?
Drop me a message anytime:
hello@lindamohamed.com