Version 1.

AI Hackathon
playbook

https://ifor.ai/
Table of contents
Introduction 4
Goals & Outcomes 5
Define the intent of the hackathon
Align goals with company profile
Clarify desired outcomes
Identify the format that fits your goals

Team & Stakeholder Alignment 7

Key roles
How to gain leadership buy-in
How to create internal excitement
Communication tips before launch

Timeline & Planning 9

4–6 weeks before
2–3 weeks before
Hackathon week
After

Theme & Challenge Design 11

How to define a good theme
Example themes
Sample challenge types

Team Formation 12
Recommended Team Size
Formation methods
Including Non-Developers

Tools, Data & AI Stack 13

Tech planning based on participant background
Tool types by function
Common tech elements

Agenda 15
Sample agenda elements
Optional additions

Mentorship & Support 15

Role of mentors
When and how to schedule mentor check-ins
Self-Guided Learning vs Structured Help

Results Submission 17
Submission format
Internal review or leadership pitch sessions
Optional: public or internal demo day
Table of contents
Judging & Recognition 18
Judging criteria examples
Judging process
Prizes & awards
Celebration ideas

Follow-Up & Impact 20

Internal survey and feedback loop
Keep the momentum
Post-event report or summary

About iForAI 21
 4
Introduction

From Experience: A Note from Ofer

“I’ve been working in AI for years - long before it became the buzzword it is today. And one thing I’ve seen

again and again is that most companies don’t struggle with knowing AI is important. They struggle with

knowing what to do about it.

That’s why I’m such a believer in AI hackathons.

Running a focused, hands-on AI hackathon is one of the most effective ways to get teams moving. It

builds confidence. It unlocks real, relevant ideas. And it gets people talking, trying, and building - instead of

just planning.

This playbook is a collection of everything we’ve learned from running AI hackathons with companies of all

sizes. Whether you’re doing this for the first time or looking to level up, I hope it gives you a useful roadmap

- and maybe even sparks some breakthroughs of your own.”

- Ofer Hermoni, Ph.D., Chief AI Officer at iForAI

What is an AI hackathon?

An AI hackathon is a short, intensive event where cross-functional teams come together

to build, test, and present ideas that use AI.

It’s not just for developers - product managers, designers, marketers, analysts, and

operators can all take part. The goal isn’t just writing code. It’s about exploring how AI can

solve real problems inside your company.

Teams usually work over a few hours or days, often with access to AI tools, internal data,

and mentorship. The result: practical prototypes, new ideas, and a big boost in AI

confidence.

Who this playbook is for

This guide is for anyone looking to run an AI hackathon inside their organization. Whether

you’re:

An innovation or transformation lead

A product or engineering manager

A learning & development owner

Or an exec sponsor looking to kickstart AI exploration

We’ll guide you through each step - from planning and setup to judging and follow-up.

This is based on our real-world experience with AI hackathons in various industries.

Goals & Outcomes 5

A successful AI hackathon starts with clear intent. First, define what success means for
your organisation. Then, consider tools or teams. That definition will shape every part of
the event - from the agenda to the judging criteria.

Define the intent of the hackathon

Your hackathon can have different purposes based on your team's AI skills and readiness.
In our experience, most hackathons fall into one or a combination of three buckets:

Raise AI awareness and excitement

Perfect for organizations early in their AI journey. These hackathons focus on giving teams
hands-on experience and building curiosity. The goal is to make AI feel approachable and real.

Explore relevant AI use cases

For teams that understand the basics and want to dive deeper. The aim is to generate ideas
that could turn into real business opportunities, even if they are still in early stages.

Prototype automation or AI-enhanced workflows

Ideal for mature teams ready to experiment with real implementation. These hackathons aim
to create demos. They improve internal processes, automate tasks, or enhance product
features.

Align goals with company profile

The nature of your organization will influence how you approach the hackathon. Being
realistic about your team’s skills helps you pick the right format, tools, and scope.

Developer-focused teams Business or non-technical Cross-functional pairing

might gain from open-ended teams

is another powerful option -

challenges. These challenges may focus on coming up with where developers collaborate

inspire them to work with real ideas. They can use no-code with non-technical

code, APIs, and data. tools, prompt engineering, or teammates. Developers bring

process design. Mentors or AI technical execution, while

specialists can provide domain experts, designers, or

guidance. operations staff contribute

user insights, real-world

challenges, and business

context.
 6
Clarify desired outcomes
Ask yourself: what do we want to walk away with?
New ideas for AI applications
Early-stage prototypes or demos
Upskilling across departments
A shift in culture and mindset around experimentation
Stronger cross-functional collaboration
Set expectations early, and communicate them clearly to participants and leadership.

Identify the format that fits your goals

Not every hackathon needs to be a 48-hour coding marathon. Depending on your intent,
you may choose a:

Short ideation session with mini-demos

1-hour or 2-hour challenge (individual format)
Perfect for quick, solo participation. Ideal for simple tasks like spotting automation chances, crafting
effective prompts, or creating a chatbot using no-code tools. Great as an intro to AI for non-technical
teams.

Half-day hackathon (≈ 5 hours)

Fast-moving, high-output session for small teams. Works especially well when using no-code or low-
code tools. Many groups can create working MVPs within this timeframe.

1-day sprint
A full-day format for teams to explore a problem, build a solution, and present it. Best for teams with
some technical or product background who can move quickly from idea to prototype.

Multi-day async
Designed for distributed or remote teams, this format gives participants 2–5 days to work in shorter,
flexible blocks. Encourages deeper exploration while staying manageable across time zones.
 7
Team & Stakeholder Alignment

Even the best-planned AI hackathon will fall flat without the right people behind it.

Success depends on early alignment across roles, clear ownership, and internal

excitement. Here’s how to make it all click.

Key roles

To run an effective AI hackathon, you’ll want buy-in and participation from across the

organization. These are the critical roles:

Executives Hackathon Lead

Give strategic support, endorse the initiative Owns the planning and coordination. Ideally

publicly, and attend key moments (like kickoff or someone who understands both AI and internal

demos). Their visible support builds legitimacy. dynamics. Think of them as the hackathon’s

“product manager.”

Cross-functional team members

Bring in people from:

Engineering
Product
Optional - Data/AI

for technical prototyping to frame business value to connect ideas to real

opportunities

Ops / Biz teams

Design / UX

to bring use cases an d pain points to help teams communicate clearly

an d craft user-centered prototypes

x
The broader the mi , the more creative and impactful the outcomes.
 8
How to gain leadership buy-in
You don’t need every exec to be an AI expert, but you do need them to believe in the
opportunity. Here’s how to get them on board:

Frame the hackathon as a low-risk way to explore AI’s potential inside the org.

Connect the goals to real business needs: productivity, customer experience, innovation, etc.

Emphasize learning outcomes: upskilling, mindset shift, visibility of internal talent.

Share success stories from similar orgs (if available).

A short kickoff or closing speech from leadership can go a long way in signaling
importance.

How to create internal excitement (pre-events, teasers, champions)

An AI hackathon should feel like something fresh and energizing - not just another
meeting. Here are a few ways to build buzz:

Pre-event teaser campaign: share the “why,” early details, and what participants can expect to gain.

Feature success stories: show how past ideas turned into real projects (or could have).

Create internal champions: recruit team leads or AI-curious employees to help promote the event
and act as early adopters.

Offer micro-incentives: limited swag, time blocks on the calendar, eye-catching graphics on Slack,
small things that show this is a real initiative.

Internal energy builds early momentum. Make it feel like something people want to join,
not something they have to do.

Communication tips before launch

Consistent messaging will help ensure strong attendance and team alignment. Here’s
what to focus on:

Be clear about the purpose: is this about experimentation, upskilling, innovation?

Share logistics early: dates, tools, expectations, prep (if any).

Remind them it’s for everyone, not just engineers.

Keep updates short, visual, and frequent in the lead-up (email, Slack, calendar events).

The more accessible and inclusive the communication feels, the broader and more
committed the participation will be.
Timeline & Planning 9

An AI hackathon may last just a few hours or a full day. However, success needs careful
planning in the weeks before the event. Here’s a simple timeline to help you prepare,
execute, and follow up.

4–6 weeks before

This phase focuses on vision and logistics.

Define goals and scope

Ensure your hackathon has a clear purpose
(see Section 2). This will guide decisions
about tools and judging.
Book the venue or set up the virtual
environment
If in-person, confirm a space that promotes
collaboration. For virtual events, prepare
necessary platforms (e.g. Zoom, Slack,
FigJam, Notion).
Finalize challenge themes
Select themes that are broad enough to
inspire but focused enough to feel
achievable. (More on this in the next
section.)

2–3 weeks before

This is your time for onboarding and building excitement.

Form teams or recruit participants

Encourage diverse roles. Make it clear that
the event is for everyone, not just engineers
(if non-engineering teams are involved).
Share AI toolkits and training
If you’re using new tools (e.g. LLMs, data
platforms), offer tutorials, demos or
recorded introductions. This helps level the
playing field.
Announce the full agenda
Inform teams when they need to be ready,
what to expect, and when demos will occur.
Clarity boosts energy.
Hackathon week
10

It’s time to put your plan into action.

Kickoff event
Welcome everyone, restate goals, and
explain the structure. Having an executive
or champion open the session adds value. Hacking sessions with mentor
support
Let teams work independently but provide
touchpoints for advice, encouragement,
Presentations and judging and technical help.
Keep presentations short (3–5 minutes)
and structured. Judging criteria should be
clear and consistent. After each
presentation allow 3-5 minutes for
questions from the judges or the audience. Celebration and

wrap-up
Name the winner, recognise all teams,
highlight top ideas, and thank participants
publicly. This shapes how people remember
the event.

After
The energy can continue even after the event ends.

Follow-up sessions or demos

Offer select teams a chance to present to
leadership, receive feedback, or explore
next steps.
Survey for feedback
Ask what worked, what didn’t, and what
they’d like next time. Keep it short and
anonymous.

A ready-to-use survey template is

included at the end of this handbook (see
Organizer Toolkit).
Track project continuation
Some ideas may evolve into pilots or
become internal tools. Capture these
stories and share them.
Theme & Challenge Design 11

​Choosing the right theme and challenge types is one of the most important parts of a
successful AI hackathon. The goal isn’t to be overly narrow, it’s to guide and inspire teams
while leaving space for creativity.

How to define a good theme

A great hackathon theme should be:

Clear Relevant Open-ended

It should be easy to It should connect to real It should allow multiple types
understand at a glance. business challenges or of ideas to emerge.
opportunities.

A good theme acts as a creative anchor. It keeps teams aligned without boxing them in.

Example themes
“AI for Customer Experience” Explore how AI can improve support, personalization, or
onboarding.

“Automate the Mundane” Encourage teams to identify repetitive work and

prototype smart automations.

“Making Our Metrics Actionable” Focus on surfacing insights from internal data and
connecting it to decision-making.
Other theme examples could include:
“From Question to Insight” “AI for Internal Workflows” “One AI Agent per Department”
 12
Sample challenge types

Here are some common challenge formats that suit various organisations and maturity
levels:

Workflow automation
Find repetitive internal tasks and use AI to cut down on manual steps or remove bottlenecks.

Product enhancement
Use AI to improve or extend customer-facing features, such as personalization,
recommendation, or customer support.

Data visualization
Turn raw data into insights using AI to automate charts, dashboards, or report summaries.

Knowledge retrieval
Build agents that can answer internal questions using documentation, CRM data, or support
archives.

You can either pre-define a few challenges or let teams frame their own within a broader
theme.

Team Formation

A great AI hackathon isn’t just about brilliant individuals, it’s about well-rounded teams
that can collaborate across skill sets. How you form those teams has a big impact on
energy, creativity, and outcomes.

Recommended Team Size

Ideal range: 3–5 people

This keeps communication fluid while giving the team enough diversity of thought and
capability to get things done.

For shorter or individual-format hackathons (like 1–2 hour challenges), it’s also effective to
let participants work solo or in pairs. This format works especially well for prompt-based
tasks or no-code chatbot building.
 13
Formation methods:

There’s no single right way to create teams, but these two approaches work well:

Self-organized Assigned teams

Let participants choose their own teams ahead Create balanced teams in advance, mixing
of time or during kickoff. This works best in functions, departments, or seniority. This is
organizations with strong informal networks or helpful if you want to encourage new
repeat participants. collaboration or ensure inclusion.

Pro tip: Share sample team templates or role suggestions (e.g. “If you have a
developer, a designer, and a PM, you’re good to go!”).

Not every team needs every role, but every team needs range.

Including Non-Developers

AI hackathons aren’t just for coders. Here’s how to make non-technical team members
successful:

Pair them with builders who can translate ideas into action.

Encourage use of no-code tools or prompt-based frameworks.

Give examples of roles they can play: prompt crafter, use case definer, presenter, user tester.
Offer lightweight templates that guide structure and keep things moving.

When everyone sees where they fit, they’re more likely to contribute, and enjoy the
process.

Tools, Data & AI Stack

Tech planning based on participant background:
Start by understanding who your participants are:

Developers Non-developers
will want flexibility, access to code environments, will thrive with no-code tools, AI interfaces, and
APIs, and robust documentation. guided templates.

Plan for both profiles and consider pairing people with different strengths together.
 14
Tool types by function

Think about the kinds of projects teams might build, and offer tools that match. Here’s a

breakdown:

Marketing tools

For content generation, campaign optimization, and audience segmentation

Engineering tools

For coding, debugging, deploying prototypes, or working with data

Design & content tools

For UI/UX prototyping, slide design, or content workflows

Internal vs external platforms

Decide if you’ll provide access to your company’s own tools, APIs, or datasets, or rely entirely on public/

third-party tools.

Common tech elements

These are the building blocks we see used most often across successful AI hackathons:

Safe environments & sandbox access A ccess to internal data or test datasets

If using internal systems, provide sandbo es or x If you can safely provide company-specific data,

dummy credentials. Make sure no one is do it. O therwise, offer sample datasets (e.g.,

touching production data or tools without customer feedback, logs, product metadata . )
guardrails.

V ibecoding platforms AP Is & prompt frame w

orks

Browser-based code environments like Replit, T eams benefit from structured tools like

Base 44 S
, tom, or Lovable make it easy to build L ang C L
hain, Prompt ayer, or Flowise for chaining

quickly without setup pain. AI capabilities or handling logic.

L arge L anguage Models (LL ) Ms

O ffer access to tools like ChatGTCP , laude, G x x x

emini, Perple ity. Provide conte t and e amples of how

they can be used ( x

e.g., prompt-based automation, te t classification, chatbot workflows . )

P ro tip: Host a short pre -v e ent session (or share a toolkit doc ) x
that e plains the tools

a vailable and how to get started. This will dramaticall y reduce friction during the e vent

itself.
Agenda 15

Every AI hackathon has its own pace. Still, a solid structure helps teams stay focused, work
together, and keep their energy up. This section lists key agenda items to help you create
your perfect flow.

Sample agenda elements

You don’t need to fill every hour, but a well-paced agenda ensures energy and clarity
throughout.

Kickoff briefing Tool training / office hours

Frame the goals, agenda, and tools. Share Give space for participants to explore new
examples to spark inspiration and clarify platforms. Schedule mentors or product owners
expectations. to be available for ad hoc support.

Regular check-ins Pitch prep

Quick standups or async updates to keep teams Block out time for teams to prepare their final
aligned and surface blockers early.
presentation. Offer a simple pitch deck or
template.

Demo presentations
Give each team 3–5 minutes to show what they built. Encourage live demos when possible, or record
them if async.

O ptional additions
M idday energy breaks
Lightning talks (5-min intros to tools or concepts)
P eer reviews or mock judging

M entorship & Support

Mentors play a key role in making an AI hackathon productive and inclusive. This is
especially true for teams new to AI or those working across different fields. But their value
depends on how clearly they’re positioned and how accessible they are throughout the
event.
 16
Role of mentors (internal or external)
You can recruit mentors from both inside and outside your organization. Their roles can
include:

Technical guides Business advisors

helping with tools, APIs, and AI workflows keeping ideas grounded in real business value

Design/UX coaches AI specialists

helping teams improve communication and user helping evaluate feasibility or guide better
flow prompts

They don’t need to be assigned to a team full-time, they can float and offer support as
needed.

When and how to schedule mentor check-ins

Plan structured touchpoints that are baked into the agenda:

Pre-scheduled time slots (e.g., 15-min per team)

Floating support windows (mentors available in breakout rooms or on Slack)

Targeted check-ins after ideation or before final presentations

Teams should know when mentors are available and how to contact them. This can be
through a mentor table, a Slack channel, or a sign-up sheet.

Self-Guided Learning vs Structured Help

Encourage a balance:

Self-guided learning: Structured help:

Teams can explore tools, data, and ideas at their Offer
basic tools like templates, example
own pace. This approach fosters ownership and prompts, or a starter kit. This helps to lower
sparks innovation. mental strain.

You want participants to feel supported but empowered. The goal is to nudge, not
babysit.
Results Submission 17

Capturing and sharing what teams build during the AI hackathon is essential, not just for
judging, but for learning, visibility, and follow-through. Keep the submission process
simple, but structured.

Submission format
Teams should submit their work in one or more of the following formats:
Pitch deck Demo video Short write-up
A simple slide deck outlining A 1–3 minute screen A brief explanation (1–2
the problem, approach, recording or walk-through of paragraphs) covering what
solution, and potential the working prototype. Tools was built, how AI was used,
impact. like Loom or Zoom recordings and what could be done next.
work great.

Pro tip: Provide a template or checklist so submissions are consistent and easy to
review.

Internal review or leadership pitch sessions

Depending on your goals, you can:
Host a private review session with hackathon organizers and mentors
Invite executives or leadership to a short pitch round (adds visibility and buy-in)
Set up peer-to-peer reviews to promote sharing and inspiration across teams
The goal isn’t only to score, it’s to celebrate effort and uncover future potential.

Optional: public or internal demo day

For higher-impact or externally-facing hackathons, consider organizing a short demo
day:
Each team presents their project to the company (or broader audience)
Add voting, shoutouts, or crowd-favorite awards to boost engagement

This gives participants recognition and shows the rest of the org what’s possible when
teams lean into AI.
Judging & Recognition 18

Judging is more than just picking winners. It’s about recognising effort, highlighting great
ideas, and supporting the values of your hackathon. Recognition fuels momentum. Here’s
how to do both effectively.

Judging criteria examples

Use 3–5 criteria to keep things focused and fair. Common ones include:
Innovation AI Integration Feasibility
Is this a new idea or a How meaningfully was AI Could this realistically be
creative approach to an used? Is the solution built implemented in your
existing problem? with or on AI, not just next to organization or product?
it?

Business Impact Presentation & Pitch

Does this solve a real pain Was the solution explained
point? Could it save time, clearly? Did the demo tell a
reduce costs, or improve compelling story?
customer experience?

You’ll find a sample scoring sheet template in the Organizer Toolkit.

Judging process
Keep it smooth and fair with a clear structure:

Panel setup
Choose 3–5 judges from a mix of functions, product, tech, business, and leadership. Consider adding
an external AI advisor if available.

Time limits
Allocate 3–5 minutes per pitch, with an optional 3–5 minutes for Q&A.

Q&A time
If time allows - let judges ask one or two clarifying questions. It helps surface team thinking and assess
depth.

Scoring system
Use a shared sheet or form with a 1–10 scale per criterion. Average the scores or use discussion to
determine winners.
 19
Prizes & awards
Recognition boosts engagement and builds momentum beyond the event. You don’t
need a huge budget to make it meaningful.

Award categories (choose 2–4)

Best U
Most Valuable Use Cas
Most Creativ
Best Technical Executio
Crowd Favorit
AI selected (provide the AI with the artifacts, and allow it to select the winner)

Tangible rewards
Swa
Gift card
Lunch with leadershi
Book
AI tools credits

Intangible rewards
Feature selected for company roadma
Public shout-out
Badge
Early access to internal tool
Mentorship follow-u
Pilot funding

Celebration ideas
Post results in company channels or newsletter
Host a short awards ceremony at the end of the event

Even small gestures of recognition can make people feel proud, seen, and eager to keep
learning.
Follow-Up & Impact 20

An AI hackathon isn’t just a moment, it can be a meaningful catalyst. But only if you follow
through. Post-event activities help you learn what worked, show participants their work
matters, and keep energy flowing into real outcomes.

Internal survey and feedback loop

After the event, send a short, focused survey to all participants and mentors. Ask
questions like:
What would you rate your level of familiarity with AI tools before the hackathon?
What would you rate your level of familiarity with AI tools after the hackathon?
What was your favorite part of the event?
What challenges did you run into?
What tools or support would you have wanted more of?
How satisfied were you with the overall event?
Would you participate again?

Also consider hosting a short retrospective with your organizing team to document what
went well and what to improve.

Keep the momentum

Don’t let great ideas die on the whiteboard. Use the hackathon as a starting point for real
change by:

Adding top ideas to your Offering project Continuing recognition

roadmap incubation support for top contributors
Identify projects that align Let teams continue Highlight top-performing
with business priorities and development with teams in company comms,
assign owners. mentorship, time, or tools, all-hands, or team meetings.
even as a 1-month pilot or Give them the stage.
internal experiment.

The best follow-up turns a fun day into a seed for transformation.
Post-event report or summary
21

Capture the event’s value in a short wrap-up:

Highlights (photos, ideas, winning teams)

Participation stats

Key learnings

Next steps or business decisions made

Send it to execs, team leads, and participants, it reinforces the ROI and invites future
engagement.

About iForAI
iForAI helps companies move from AI interest to real-world impact.

We specialize in:

Upskilling: lectures, workshops, and internal hackathons

Strategy: use case identification, roadmap building

Execution: solution delivery or mentoring your internal teams

Whether ou’re ust getting started or loo ing to accelerate existing efforts, we wor side-
y j k k

-side with our team to ma e it happen.

by y k

Whe r to i
e f nd u s

G et n tou
i ch F o o us on so a me a
ll w ci l di More a out us
b

hi ifor.ai
@ iForAI https: ifor.ai
// /

Ofer ermoni h. .
H P D