Our First Company-Wide ChatGPT Hackathon After Full Rollout

Happy New Year. Okada from Zenken’s AI division here.

On January 7, 2025 — only the second working day of the year — more than 40 people gathered for our ChatGPT hackathon, including executives and division heads from ten different business units.

In about three valuable hours, each team enjoyably learned the GPTs feature and built a working prototype that solves something practical.

1. Brainstorming for new business lines
2. Streamlining email correspondence with overseas partners
3. Drafting job postings
4. Training Smith — an English-learning advisor
5. A funnel-analysis assistant for job-seeker behavior
6. Initial troubleshooting for IT devices
7. Project manager for the “shareholder Soka” project
8. Payroll calculation

All eight teams produced something genuinely unique.

At the end, everyone voted for the GPT they thought was best. Many were close calls, but the runaway winner was Initial troubleshooting for IT devices.

The problem it solves For laptop and smartphone trouble, it walks the user through what they can resolve on their own, and prepares an escalation message for cases that need IT support.

The benefits Basic problems get cleared without help, and when escalation is needed, the message clearly conveys what is happening — which cuts response time. Users don’t have to wait. It also builds the user’s IT skills: by going through initial troubleshooting themselves, they learn over time.

The GPT walks the user step by step, narrowing down the issue without requiring technical expertise. If three steps don’t resolve it, or if administrative privileges are required, it switches into escalation mode.

I think it won decisively because of one thing: it tackles the invisible labor that doesn’t show up in any dashboard. Both the IT team’s time and the user’s waiting time get cut, and the savings on both sides are concrete and easy to picture.

This was also an event where every section of the corporate group joined in — work that “quietly supports the business from behind the scenes” suddenly became visible to a much larger audience. There was a lot of mutual learning in the room.

We started preparations at the end of 2024 with help from the OpenAI Japan team, and I’m glad we were able to pull it off. Through this event, many people surfaced challenges, learned new ways to use the tool, and walked away with insights they’ll bring back to their work.

We’ll keep pushing — not just within specific departments, but across every part of Zenken — and the AI division will stay focused on raising the level of usage company-wide.