Gen AI has dominated the conversation, yet most organisations remain unsure of where to begin. AWS recognised a clear opportunity: create a space where people could build, test, and learn, guided by real tools, real experts, and real use cases.
So we designed an environment that felt open and intuitive, built around hands-on discovery. No long presentations. No passive audiences. Just a place where Amazon Bedrock, Titan, CodeWhisperer, and Amazon Q came alive through exploration. Everything, from the spatial layout to the programming, was shaped to make the journey from "what is possible" to "let's build it" feel natural.
2Moons was reimagined as a modular, future-forward workspace. Demo pods, experimentation tables, adaptive lighting, and AI-generated art shaped a setting that was both functional and alive. It carried AWS's clarity without slipping into minimalism for its own sake. The design stayed out of the way, allowing the work to take centre stage.
We didn't optimise for volume. We prioritised intent. Each day was built around a specific cohort: enterprise leaders, developers, product teams, ISVs, and AI-native startups. This structure ensured that sessions were relevant, conversations were sharper, and every attendee found room to explore their own context.
Our teams anchored the Loft end-to-end: from running technical sessions and live demos to producing content and managing real-time storytelling. Social amplification followed naturally through AWS leaders and community voices, giving the program reach without forcing a narrative. The experience moved with precision, but never felt rigid.
The impact extended beyond metrics. The Loft offered clarity in a space often defined by noise. People left not with abstract possibilities, but with workable next steps.
The AWS Gen AI Loft created a setting where building felt accessible, guided, and worth the time invested. By structuring the experience around hands-on value and intentional curation, the Loft set a practical standard for how Gen AI should be introduced, not as a concept, but as a capability.