AWS Builder Center
AWSKRUG AI Engineering: OpenClaw on AWS Workshop Recap

AWSKRUG AI Engineering: OpenClaw on AWS Workshop Recap

Building personal AI agents on AWS with OpenClaw : a recap of AWSKRUG AI Engineering's August hands-on workshop.

Why This Topic?

Interest in AI agents has never been higher. But the moment you decide to actually build one, the story changes. Picking a framework, setting up the environment, wiring up tools: most people run out of steam somewhere along the way. There's plenty of curiosity, but surprisingly few chances to get hands on the keyboard.
That's why our August AWSKRUG AI Engineering meetup ran a hands-on workshop building and using OpenClaw, an open-source AI agent platform, on AWS. This event was supported through the AWS community meetup support program, which we applied for and received funding from.
The AWSKRUG AI Engineering meetup focuses on the practical problems that developers face in the generative AI era, including prompt engineering, RAG techniques, and generative AI development tools that cloud and software engineers should know about.

What We Built

The goal was to give participants an end-to-end experience of building a personal AI agent. Not a session about what AI agents are in the abstract, but one where you connect real tools and get an agent running.
1. Email integration with AgentMail
The first tool connection. Email is something we touch every single day, and the moment an agent can read and send it for you, the usefulness becomes immediately tangible.
2. Calendar integration with CalDAV (Radicale)
Next, we connected a calendar so agents could see and manage schedules.
3. Document generation skills
Participants then had their agents produce PDF and PowerPoint files, turning agent output into something you can actually hand off.
4. Building AI agents (Essentials)
  • Daily Brief Agent, which summarizes your day
  • Productivity Agent, which assists with everyday work
With the tools in place, this is where the pieces came together into something that behaves like an assistant.
5. Building specialized agents (Applied)
Finally, participants built agents tailored to specific domains: Research, Business and CloudOps.




Why No Coding Experience Was Needed

The entire workshop runs through the OpenClaw web UI, so participants completed every exercise without writing a single line of code.
That meant not just developers, but also product managers, students, and participants from non-engineering backgrounds could follow along hands-on. AI agents are often treated as developer-only territory, and lowering that barrier felt like a meaningful thing to do.

Scenes From the Day

Using the budget, we prepared custom treats(macaron) featuring the OpenClaw and AWSKRUG AI Engineering meetup characters. A small touch, but participants took plenty of photos, and it definitely helped create the kind of atmosphere that community events are known for.




Lessons From Running It

Environment setup took longer than expected.
With around 70 participants each preparing their own laptop, the setup time ate into the hands-on portion. In a two-hour workshop, those early minutes are felt keenly.
For future events, we plan to share the setup guide 15 minutes before the session starts, so participants can get their environments ready before we begin.
A bit more explanation of OpenClaw itself would have helped.
Because we moved quickly through the hands-on portions, some participants wanted a clearer picture of how things work under the hood. Running the exercises matters, but so does understanding the principles behind them: without that, following the steps can feel mechanical.
For the next event, we're considering putting together supplementary material on the core concepts and architecture and sharing it with participants in advance, so they arrive with the conceptual groundwork already in place and can spend the session connecting it to what they're building.

Wrapping Up

AI agents are moving from "good to know about" to "you should actually try one." I hope this workshop gave participants that first hands-on taste.
A Huge thanks to AWS for supporting this meetup. It's what let us prepare something special for participants on the day, like the custom macaron, which added a lot to the atmosphere! 🎁
Also thank you to the AWSKRUG AI Engineering organizers who ran this event with me. From planning to running the day itself, none of it happens without them! 😎

The AWSKRUG AI Engineering meetup runs monthly, covering practical topics from the generative AI development landscape. If you're interested, join us on  AWSKRUG Meetup  and  Slack .
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article