02
days
15
teams
10
per team
01
challenge
THE AGENDA
Two days, minute by minute
From the first keynote to the final demo.
Tuesday 27 October 2026
Wednesday 28 October 2026
DAILY HIGHLIGHTS
The story, day by day
Team of the day, volunteer spotlights and photos — added at the end of each day.
Day 1 · Tue 27 Oct
TEAM OF THE DAY
Bright minds are in the room, speakers are setting the stage, and the ideas are already flowing.
SPEAKERS, PANEL & EXPERTS
The people behind the ideas
Business, technology and human expertise, together.
Day 1 opens with executive keynotes and a session on data and human behaviour. Profiles marked to follow are added as they are confirmed.
Day 1, 10:35 to 10:50, moderated by Mohamed BinRuken.
The judging panel · Day 2. Scoring framework proposed and pending approval.
Mentors from the UAE National AI Program and Nokia help the teams test ideas, unblock technical issues and sharpen their demos.
What teams will build
THE CHALLENGE · REVEAL COUNTDOWN
One challenge.
Fifteen possibilities.
The brief and challenge Q&A are locked until the Day 1 challenge introduction.
27 October 2026 · 10:50 Dubai time (UTC+4)
00
days
15
hours
46
minutes
40
seconds
UAE CODES DAY · AI CHALLENGE
Build an AI conversation simulator
Help Emiratis practise the conversations that matter — with an AI character that listens, adapts and helps them improve.
One challenge. Fifteen teams. Real possibilities for the UAE workplace.
YOUR BUILD BRIEF
Questions about the challenge
Pick a topic to see the answers. 19 questions, five topics.
Build an AI conversation simulator that helps Emiratis practise challenging professional situations across different industries. Users choose a situation, practise a conversation with an AI character, receive feedback on how they did, then retry with a different approach and compare the two attempts.
Any conversation people often find difficult at work, such as asking for a promotion, giving difficult feedback, handling an upset customer, negotiating with a supplier, pitching to an investor or disagreeing with a manager. Choose situations that feel realistic and relevant to the UAE workplace.
The brief names three groups: early-career employees, experienced professionals and business owners. Your prototype should include realistic scenarios for each.
Yes. Your scenarios should span more than one industry, such as banking, healthcare, retail, government or technology. A well-chosen mix is enough; you do not need to cover every industry.
1) Choose a situation. 2) Practise the conversation with an AI character whose responses adapt to what the user says. 3) Receive feedback on clarity, listening, empathy and assertiveness, supported by examples from the conversation. 4) Retry with a different approach and compare the two attempts.
The AI character should react to the user’s actual words rather than follow a fixed script. Two users taking different approaches should end up in noticeably different conversations.
Specific and evidence-based. Quote what the user actually said, explain why it worked or did not, and suggest how to improve. Cover all four skills: clarity, listening, empathy and assertiveness.
After a retry, show the user what changed: how their feedback differs between attempts, and which approach worked better and why.
It must show one full journey working end to end: choose a situation, hold the conversation, receive feedback, retry and compare. A working journey matters more than polished screens.
Yes. Laptops are collected at registration so the technical team can install the required AI tool, and returned before coding begins.
Teams build with approved tools. Codex may be used if it is available and approved at that time. The approved tools list is to be confirmed. Ask a mentor or the technology support team if you are unsure.
Each team is assigned a sub-prompt when teams are announced on Day 1, narrowing its focus within the main challenge.
Mentors and technology support are available throughout the build sprints to test ideas, unblock technical issues and sharpen your demo.
No. Use fictional scenarios, characters and sample conversations only. Do not enter any real personal, customer or bank data.
Keep feedback constructive and respectful; make it clear the user is talking to an AI; avoid stereotypes in characters and scenarios; and consider how practice data would be stored and protected. Teams complete a responsible AI checklist as part of the evidence pack.
Your working demo and final PowerPoint slides, collected by the volunteers before the presentations, together with the responsible AI checklist. The submission deadline is to be confirmed.
The proposed framework has seven criteria totalling 100 marks and is pending management review and approval. View the proposed criteria. The judges agree the 1st, 2nd and 3rd place winners.
The problem you are solving and who it is for; a live demo of the full journey; and your key design, data and responsible AI decisions.
At the closing ceremony on Wednesday 28 October.
Who’s taking part
Speakers, judges and mentors join from across the UAE, alongside the 15 teams.
Emirates NBD
Speakers, judges, mentors and your master of ceremonies
Ministry of AI
Minister of State for Artificial Intelligence Office
Emirates NBD on Social
Build an AI conversation simulator
Build an AI conversation simulator that helps Emiratis practise challenging professional situations across different industries.
The user selects a situation and practises a conversation with an AI character whose responses adapt to what they say. Provide feedback on clarity, listening, empathy and assertiveness, supported by examples from the conversation.
Allow the user to retry with a different approach and compare the two attempts. Include realistic scenarios for early-career employees, experienced professionals and business owners.
Questions about the challenge
Build an AI conversation simulator that helps Emiratis practise challenging professional situations across different industries. Users choose a situation, practise a conversation with an AI character, receive feedback on how they did, then retry with a different approach and compare the two attempts.
Any conversation people often find difficult at work, such as asking for a promotion, giving difficult feedback, handling an upset customer, negotiating with a supplier, pitching to an investor or disagreeing with a manager. Choose situations that feel realistic and relevant to the UAE workplace.
The brief names three groups: early-career employees, experienced professionals and business owners. Your prototype should include realistic scenarios for each.
Yes. Your scenarios should span more than one industry, such as banking, healthcare, retail, government or technology. A well-chosen mix is enough; you do not need to cover every industry.
1) Choose a situation. 2) Practise the conversation with an AI character whose responses adapt to what the user says. 3) Receive feedback on clarity, listening, empathy and assertiveness, supported by examples from the conversation. 4) Retry with a different approach and compare the two attempts.
The AI character should react to the user’s actual words rather than follow a fixed script. Two users taking different approaches should end up in noticeably different conversations.
Specific and evidence-based. Quote what the user actually said, explain why it worked or did not, and suggest how to improve. Cover all four skills: clarity, listening, empathy and assertiveness.
After a retry, show the user what changed: how their feedback differs between attempts, and which approach worked better and why.
It must show one full journey working end to end: choose a situation, hold the conversation, receive feedback, retry and compare. A working journey matters more than polished screens.
Yes. Laptops are collected at registration so the technical team can install the required AI tool, and returned before coding begins.
Teams build with approved tools. Codex may be used if it is available and approved at that time. The approved tools list is to be confirmed. Ask a mentor or the technology support team if you are unsure.
Each team is assigned a sub-prompt when teams are announced on Day 1, narrowing its focus within the main challenge.
Mentors and technology support are available throughout the build sprints to test ideas, unblock technical issues and sharpen your demo.
No. Use fictional scenarios, characters and sample conversations only. Do not enter any real personal, customer or bank data.
Keep feedback constructive and respectful; make it clear the user is talking to an AI; avoid stereotypes in characters and scenarios; and consider how practice data would be stored and protected. Teams complete a responsible AI checklist as part of the evidence pack.
Your working demo and final PowerPoint slides, collected by the volunteers before the presentations, together with the responsible AI checklist. The submission deadline is to be confirmed.
The proposed framework has seven criteria totalling 100 marks and is pending management review and approval. View the proposed criteria. The judges agree the 1st, 2nd and 3rd place winners.
The problem you are solving and who it is for; a live demo of the full journey; and your key design, data and responsible AI decisions.
At the closing ceremony on Wednesday 28 October.
