From Ideas to Action: Highlights from Our AI Agents Hackathon

Unlike traditional chatbots, which simply respond to commands, AI agents can integrate with tools, make decisions in context, and take meaningful action. In short, they work with you — almost like a colleague. That vision inspired dozens of ideas, prototypes, and late-night breakthroughs during our recent hackathon.

On 16–17 September, more than half of RealityMine came together for our 6th internal hackathon: two fast-paced days of creativity, collaboration, and experimentation. With 39 participants across 14 cross-functional teams, the spotlight was firmly on one of the most exciting frontiers in technology: AI agents.

Themes that Emerged

Across the 14 projects, two key themes stood out:

1. Contextual summarisation pipelines
Several teams explored ways to pull scattered information from multiple sources, add context, and generate outputs that people can act on. Examples included:

• Competitor analysis automatically feeding into Confluence updates.

• Alerts and customer communications triggered by relevant industry news.

• On-call assistance agents providing diagnostic context in the middle of the night.

2. Knowledge-aware workflow assistants
Other teams tackled internal processes, using AI to combine knowledge bases, task-specific inputs, and resources to reduce manual work. Projects ranged from:

• AI-assisted job creation and validation in data engineering.

• Cross-checking client statements of work when setting up research panels.

What Happens Next?

Hackathons at RealityMine aren’t just for show — they’re a launchpad. Several projects are already being taken forward:

• Competitor analysis automation is now a protected Friday project for further development.

• Knowledge-aware job validation is moving toward integration with AWS resources.

These experiments don’t just create potential tools — they also help us learn how to run effective AI projects, improve prompting, and design better context-aware workflows.

Why It Matters

AI isn’t a side project for RealityMine. It’s core to our strategy — both in how we build products and how we work internally. Hackathons like this give us space to explore, learn, and test what AI can and should do for us.

As I see it, the value isn’t just in the projects themselves, but in what we learn together as a business — how to use AI more effectively, how to prompt better, and how to design agents that genuinely help.

This hackathon was just the beginning. The projects, learnings, and excitement it sparked will continue to shape how we innovate with AI at RealityMine.

About the Author

Graham Dean is Chief AI Officer at RealityMine, where he leads the responsible adoption of AI across products, processes, and teams. He joined as the company’s first engineer in 2012 and served as CTO for seven years before shifting focus to AI strategy, ethics, and governance.

‍