Practical AI, built around how your business actually works.

AI hackathons and AI workshops, explained

What is an AI hackathon?

An AI hackathon is a single working day in which a business team takes one real problem and builds a working AI prototype for it, using the team’s own data and systems rather than a demo environment. It is not a coding competition, and it does not require developers. The morning is spent framing the problem and narrowing it to something that can realistically be built, and the afternoon is spent building and testing with the people who would actually use the result. A corporate AI hackathon ends with three things: something that works, a record of how it was built, and a decision on whether to take it into production.

What is an AI workshop?

An AI workshop is a facilitated half-day session in which a business team works out where artificial intelligence would actually create value in its own processes, rather than learning about AI in the abstract. A good AI workshop produces decisions, not slides. Participants map candidate use cases across sales, service, marketing, operations, and administration, score each one on business value, effort, and data readiness, and leave with a prioritised shortlist and a clear recommendation on what to build first. We recommend starting here rather than with a hackathon. It is the smaller commitment, it tells you whether a build day is justified at all, and it means that when you do book one, the whole day goes on a problem worth solving.

Sirocco runs AI workshops and AI hackathons for companies in Stockholm and across the Nordics, as an independent partner. Most engagements start with the half-day workshop.

Three ways to start

Each format stands on its own. Most teams start with the workshop and book the hackathon from there.

AI hackathon and AI workshop FAQ

What is the difference between an AI hackathon and an AI workshop?

An AI workshop is about deciding what to build. An AI hackathon is about building it. In a half-day workshop, a team maps candidate use cases, scores them on value, effort, and data readiness, and leaves with a prioritised shortlist. In a one-day AI hackathon, that shortlist becomes a working prototype running on the company’s own data, tested by the people who would use it. Most teams run the workshop first, because it is the cheapest way to find out whether a hackathon is worth booking and which use case deserves the day.

What happens in a half-day AI workshop?

A half-day AI workshop runs for about three hours with the leadership team and the people who own the processes. It opens with a concrete view of what AI can and cannot do in your sector, then moves into mapping your own processes to find candidate use cases across sales, service, marketing, operations, and administration. Each candidate is scored on business value, effort, and data readiness. The workshop ends with a prioritised shortlist, an honest view of how ready your data is, and a recommendation on what to build first and in which format.

How long does an AI hackathon take?

A corporate AI hackathon runs for one day, typically 09:00 to 16:00. The morning is spent framing the problem and cutting it down to a slice that can realistically be built. The afternoon is spent building and testing. The last fifteen minutes are spent deciding what happens next. Half-day formats work well for discovery but rarely produce anything usable, because building and testing a prototype on real data takes most of an afternoon. Taking the prototype into production for a live team typically needs a further two to six weeks.

How much does an AI hackathon cost?

An AI hackathon is priced per session, based on the format, the number of participants, and how much preparation the use case needs. Sirocco confirms a fixed price and a written scope before anything is booked, so there is no open-ended engagement to approve. A half-day discovery workshop is the smallest commitment and the usual starting point, and it is often the cheapest way to find out whether a full hackathon day is justified. Book a meeting and we will give you a price for your specific situation.

How many people should take part in an AI hackathon?

Six to twelve people works best. Fewer than six and you lose the range of perspectives that makes the prioritisation honest. More than twelve and the building stops being hands-on. The room should include the people who own the process, someone who knows where the data actually lives, and at least one person with the authority to decide what happens after the day ends. Bringing that decision-maker is the single biggest predictor of whether the prototype survives the week after the hackathon.

Where do you run AI hackathons?

On site at your own office, or at Sirocco’s Stockholm office. Sessions run in English or Swedish. Sirocco is headquartered in Stockholm, Sweden, with teams in Barcelona, Turin, Dubai, and New York, so hackathons and workshops can be delivered across Europe, the Middle East, and the US. In-person works better for the hackathon format, because the value comes from having the process owners and the decision-maker in the same room for the full day.

Do we need our data in order first?

No. One of the most useful outputs of a discovery workshop is an honest view of how ready your data actually is, which is usually better than teams fear in some areas and worse in others. For a hackathon we work in a sandbox or a copy, so nothing in production is at risk, and we scope the prototype to data you already have rather than data you wish you had. If the real work is data preparation rather than AI, we will say so, because that is cheaper to hear in a workshop than after a project has started.

Which AI platform will you recommend?

Whichever one fits, including options we do not sell. Sirocco is an independent, vendor-agnostic partner, so the recommendation is not funded by a platform vendor. In practice the answer is often the AI already included in a system you own, such as Salesforce Agentforce, Microsoft Copilot, or HubSpot Breeze, because that avoids a new contract and keeps your data where your governance already reaches. Sometimes a lighter tool outside the CRM is the better answer, and sometimes the right answer is to improve a process before adding AI to it.

Do we have to start with our CRM?

No. Hackathons and workshops cover the whole business, including finance, operations, HR, and internal knowledge work. CRM does come up often, because that is where customer data already lives, where the effect on revenue is easiest to measure, and where Sirocco has the deepest expertise. But plenty of strong first use cases sit elsewhere, such as document handling, tender screening, reporting, or an assistant over internal documentation. We start from where the work is painful, not from a particular system.