Practical guides

AI hackathon in companies. The IAkhathon format to build real agents in 1 to 2 days

AI hackathon in companies. The IAkhathon format to build real agents in 1 to 2 days

An AI hackathon in a company brings teams together around a simple challenge. Over one to two days, each group starts from a daily irritant and leaves with a working AI agent, built in natural language, without writing a single line of code.

This intensive format has a name, the IAkhathon. It turns the team into AI explorers, because AI is learned by taking on challenges.

What is an AI hackathon in a company?

An AI hackathon is an intensive one to two-day format where teams build real AI agents on their business use cases, guided by facilitators. No development prerequisite. Participants start from a daily irritant and leave with a working tool.

The nuance matters. A developer hackathon produces code while a business AI hackathon produces usage.

The audience goes beyond the IT department alone. It concerns innovation managers, CIOs, CHROs and AI transformation directors. All share the same imperative. Collective adoption must be guaranteed for every team, and nothing triggers the desire to learn like a playful collective experience.

Why an AI hackathon triggers adoption

A hackathon pulls three levers at once, action, ownership and collective motivation.

  • Taking action with guidance, without fearing to fail alone in front of the tool
  • Leaving with your own AI agent at the end of the workshop, rather than a list of best practices
  • Boosting collective motivation, because challenges are played in teams and the debrief happens in front of everyone
  • Leaving with a certificate of achievement and a digital badge to share on LinkedIn

The last point counts more than it seems. A badge shared publicly commits the participant to keep using what they built.

The three formats, from 2 hours to a full day

The IAkhathon comes in three durations depending on the goal. The flagship format spans 2 days with a final debrief in front of the executive committee.

Short format, 2 hours, for a quick discovery

After 20 minutes of generative AI acculturation, the facilitator launches 4 challenges of 20 minutes on predefined use cases. Facilitators guide the participants. The final challenge is to create the agent of their choice for the most advanced, with a 15-minute debrief.

Medium format, half a day, for team building

Start with 30 minutes of generative AI acculturation, then give room to creativity with use cases of their choosing. Close with a 15-minute debrief per project.

Long format, one day, for POC and prototyping

Start with 2 hours of in-depth training on generative AI, then let the teams develop their use case freely. End with a 30-minute debrief per project to celebrate the creations.

Which format to choose

For a first contact with AI, the short format is enough. To bond a team and launch a dynamic, the medium format does the job. To prototype real use cases that will live after the event, the long format or the 2-day IAkhathon are the right call.

Facilitation, the key to success

The format works from 8 people and scales to several dozens of participants split into teams. The ratio that makes the difference remains one facilitator for 12 participants, with close and personalized support. The facilitation team mixes AI experts and change specialists, which welcomes every level of expertise.

On the logistics side, three prerequisites are enough. A space with suitable chairs and tables. Every participant brings their own laptop. Presentation material for the debriefs. The rest is settled upstream.

The business domains that work best

In an IAkhathon, use cases come from the business, not from the technology. The most frequent domains are accounting, creative work, dev, marketing, finance and product. The participants’ personal topics also have their place, because they create desire and memorization.

Do you need to know how to code to take part?

No. The format is designed for business teams without technical skills. Facilitators guide each step, from framing the use case to building the agent in natural language.

Certification and OPCO funding

An AI hackathon run by a Qualiopi-certified training provider is eligible for OPCO and CPF funding for companies financing their teams’ skills. An existing training budget can thus cover an operation that would otherwise look like simple team building.

What participants take away

Each participant leaves with their own working AI agent, a certificate of achievement and a digital badge to share on LinkedIn. The agents built remain usable the next day in daily work.

The format also suits public events. Meetups mixing demonstration and live building follow the same logic, everyone leaving with an agent built in natural language.

Organizing your IAkhathon

For a first AI hackathon, frame the format around a real business irritant and keep teams on use cases they can pursue after the event. That is the condition for the collective experience to turn into durable adoption.

References

Our DNA

Arkange works with leadership teams, business functions and partners who want to move from experimentation to field adoption.

Client references

Lisi Group
Igensia Education
Audit Expertises Consultant
Pimenko
Hexagone Stratégie
Neptune Internet Services
Société d’Exploitation de la Tour Eiffel
BforBank

Service partners

  • Cursor
  • Dust
  • Qualiopi. Quality certification for training actions

A first conversation, no commitment

Ready to transform your company with AI?

We clarify your context, your priority pain points and the shortest path to measurable results.