We preloaded the challenge: ten app ideas designed to improve the Gapstars employee experience. From car park and shuttle reservation systems to a CV builder that could help young engineers land their first job, each team chose one concept to bring to life.<\/p>\n
Teams were free to form across partner and internal squads, and the only rules were to collaborate, experiment with AI, and aim for a functional MVP by the end of the day.<\/p>\n
<\/p>\n
What They Built<\/strong><\/h2>\nHot Seat Booking \u2013 Team “No Humans Allowed”<\/strong><\/h3>\n“It looked like an airplane booking system… we\u2019re proud of that.”<\/em><\/p>\nThis team created a smooth seat reservation system for shared office space, complete with filters like window seats, floor layouts, and even a vision for an AI-powered seat recommender based on personal preferences.<\/p>\n
What went well:<\/h4>\n\n- Clean, interactive UI<\/li>\n
- Clear UX flow and logical architecture<\/li>\n<\/ul>\n
Where they struggled:<\/h4>\n\n- AI didn\u2019t grasp responsive layout tweaks<\/li>\n
- Animations required several retries<\/li>\n
- Platform credits burned fast trying to fix spacing issues<\/li>\n<\/ul>\n
<\/p>\n
Their frontend experience<\/strong> made a real difference, knowing how a React app should behave helped them prompt smarter and fix faster.<\/p>\n <\/p>\n
StarPlanner \u2013 Team “FusionFocus”<\/strong><\/h3>\n“Very appealing UX… but we ran into connection issues.”<\/em><\/p>\nStarPlanner tackled internal event RSVPs with a clean design and persistent backend logic planned via Supabase. But just as they hit their stride, the AI tool went down for 20+ minutes. Just a good old-fashioned reminder that yes, even AI tools crash.<\/em><\/p>\n <\/p>\n
![\"\"](\"https:\/\/tech.gapstars.net\/wp-content\/uploads\/2025\/07\/AI-Hackathon-Team-300x200.webp\")
<\/p>\n
<\/h2>\nWhat We Learned<\/strong><\/h2>\nAcross all teams and even our own prompt logs, a few clear patterns emerged, not just about AI, but about the way our engineers think.<\/p>\n
<\/p>\n
Key Prompting Strategies That Worked<\/h4>\n
<\/p>\n
1. Step-by-Step Prompting > Big-Bang Instructions<\/strong><\/p>\nTeams that prompted in modular stages got better results. Instead of asking the AI to do everything at once, they broke it down:<\/p>\n
<\/p>\n
Create a table layout for shuttle booking with time slots
\n\u2192 Add login using email only
\n\u2192 Connect to Supabase project \u2018shuttle-prod\u2019 and fetch bookings<\/div>\n <\/p>\n
2. Contextual Framing Improved Results<\/b><\/p>\n
Prompts with real design and use case context worked significantly better than generic instructions.<\/p>\n
<\/p>\n
\u274c \u201cAdd filters\u201d
\n\u2705 \u201cAdd filters for seat location (window\/aisle), floor level, and availability on hover\u201d
\n\u2705 \u201cUse Gapstars brand colors with a modern card-based layout\u201d<\/div>\n <\/p>\n
This made outputs more usable, branded, and aligned with real-world expectations.<\/p>\n
<\/p>\n
3. Using Real Naming Conventions Helped the AI Stay Grounded<\/strong><\/p>\nPrompts that referenced real Supabase tables, project names, and component labels reduced AI hallucinations.<\/p>\n
<\/p>\n
\u2705 \u201cUse Supabase project \u2018office-zen-booking\u2019 and table \u2018massages\u2019 to retrieve availability by date\u201d<\/div>\n <\/p>\n
When vague table names like test1<\/code> or demoTable<\/code> were used, the AI often defaulted to fake logic or broken code.<\/p>\n <\/p>\n
4. Tone & Clarity Mattered<\/b><\/p>\n
Prompts structured like Jira tickets or Figma specs produced better results than chatty or ambiguous commands.<\/p>\n
<\/p>\n
\u274c \u201cCan you make this prettier and modern-looking?\u201d
\n\u2705 \u201cApply a minimal design with rounded corners, Gapstars Orange as primary CTA, and condensed spacing on mobile.\u201d<\/div>\n <\/p>\n
Clear language = cleaner code.<\/p>\n
<\/p>\n
5. Versioning Prompts Saved Time<\/strong><\/p>\nSeveral teams labeled their iterations:<\/p>\n
<\/p>\n
\u201cTry V2 of homepage with fewer filters\u201d
\n\u201cRevert to V3 and reduce card padding\u201d<\/div>\n <\/p>\n
This \u201cprompt hygiene\u201d helped with tracking changes, reusing effective blocks, and avoiding repetitive errors, especially when the project grew past 10+ steps.<\/p>\n
<\/p>\n
A Moment with Our Partners<\/strong><\/h2>\nAdding to the day\u2019s energy, we were thrilled to host Harver, one of our longest-standing partners, during the Hackathon. With product and engineering leaders joining in, it was a chance for them to see how Gapstars engineers collaborate, build, and troubleshoot in real time.<\/p>\n
It was less about the outputs, and more about watching engineering thinking under pressure something you can\u2019t really teach or fake.<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/tech.gapstars.net\/wp-content\/uploads\/2025\/07\/AIHackathon_Article-Image-300x200.webp\")
<\/p>\n
<\/h2>\nThe Takeaway<\/strong><\/h2>\nThis hackathon wasn\u2019t about building perfect apps. It was about <\/span>testing ideas quickly<\/b>, <\/span>prompting clearly<\/b>, and <\/span>building creatively<\/b> under real-world constraints.<\/span><\/p>\nYes, AI tools helped. But the real advantage came from the people behind the prompts:<\/span><\/p>\n\n- The ones who scoped well.<\/span><\/li>\n
- The ones who debugged fast.<\/span><\/li>\n
- The ones who iterated with intent.<\/span><\/li>\n<\/ul>\n
<\/p>\n
Because at the end of the day, the quality of the build still depends on one thing: <\/span>Not how good the AI is, but how good YOU ARE at telling it what to do. <\/b>And if this hackathon was anything to go by, our teams are learning to do just that. We\u2019re not just building teams. We\u2019re building teams that innovate with intent, adapt with speed, and scale with purpose.<\/span><\/code><\/p>\n","protected":false},"excerpt":{"rendered":"The secret wasn\u2019t the AI. It was our engineers how they framed problems, gave clear prompts, and iterated with intent. That\u2019s where the magic happened.<\/p>\n","protected":false},"featured_media":9476,"template":"","meta":{"_acf_changed":false,"content-type":""},"resource-category":[77],"class_list":["post-9472","resource","type-resource","status-publish","has-post-thumbnail","hentry","resource-category-blog"],"acf":[],"_links":{"self":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource\/9472","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource"}],"about":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/types\/resource"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/media\/9476"}],"wp:attachment":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/media?parent=9472"}],"wp:term":[{"taxonomy":"resource-category","embeddable":true,"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource-category?post=9472"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
“It looked like an airplane booking system… we\u2019re proud of that.”<\/em><\/p>\n This team created a smooth seat reservation system for shared office space, complete with filters like window seats, floor layouts, and even a vision for an AI-powered seat recommender based on personal preferences.<\/p>\n <\/p>\n Their frontend experience<\/strong> made a real difference, knowing how a React app should behave helped them prompt smarter and fix faster.<\/p>\n <\/p>\n “Very appealing UX… but we ran into connection issues.”<\/em><\/p>\n StarPlanner tackled internal event RSVPs with a clean design and persistent backend logic planned via Supabase. But just as they hit their stride, the AI tool went down for 20+ minutes. Just a good old-fashioned reminder that yes, even AI tools crash.<\/em><\/p>\n <\/p>\n Across all teams and even our own prompt logs, a few clear patterns emerged, not just about AI, but about the way our engineers think.<\/p>\n <\/p>\n <\/p>\n 1. Step-by-Step Prompting > Big-Bang Instructions<\/strong><\/p>\n Teams that prompted in modular stages got better results. Instead of asking the AI to do everything at once, they broke it down:<\/p>\n <\/p>\n <\/p>\n 2. Contextual Framing Improved Results<\/b><\/p>\n Prompts with real design and use case context worked significantly better than generic instructions.<\/p>\n <\/p>\n <\/p>\n This made outputs more usable, branded, and aligned with real-world expectations.<\/p>\n <\/p>\n 3. Using Real Naming Conventions Helped the AI Stay Grounded<\/strong><\/p>\n Prompts that referenced real Supabase tables, project names, and component labels reduced AI hallucinations.<\/p>\n <\/p>\n <\/p>\n When vague table names like <\/p>\n 4. Tone & Clarity Mattered<\/b><\/p>\n Prompts structured like Jira tickets or Figma specs produced better results than chatty or ambiguous commands.<\/p>\n <\/p>\n <\/p>\n Clear language = cleaner code.<\/p>\n <\/p>\n 5. Versioning Prompts Saved Time<\/strong><\/p>\n Several teams labeled their iterations:<\/p>\n <\/p>\n <\/p>\n This \u201cprompt hygiene\u201d helped with tracking changes, reusing effective blocks, and avoiding repetitive errors, especially when the project grew past 10+ steps.<\/p>\n <\/p>\n Adding to the day\u2019s energy, we were thrilled to host Harver, one of our longest-standing partners, during the Hackathon. With product and engineering leaders joining in, it was a chance for them to see how Gapstars engineers collaborate, build, and troubleshoot in real time.<\/p>\n It was less about the outputs, and more about watching engineering thinking under pressure something you can\u2019t really teach or fake.<\/p>\n <\/p>\n This hackathon wasn\u2019t about building perfect apps. It was about <\/span>testing ideas quickly<\/b>, <\/span>prompting clearly<\/b>, and <\/span>building creatively<\/b> under real-world constraints.<\/span><\/p>\n Yes, AI tools helped. But the real advantage came from the people behind the prompts:<\/span><\/p>\n <\/p>\n Because at the end of the day, the quality of the build still depends on one thing: <\/span>Not how good the AI is, but how good YOU ARE at telling it what to do. <\/b>And if this hackathon was anything to go by, our teams are learning to do just that. We\u2019re not just building teams. We\u2019re building teams that innovate with intent, adapt with speed, and scale with purpose.<\/span> The secret wasn\u2019t the AI. It was our engineers how they framed problems, gave clear prompts, and iterated with intent. That\u2019s where the magic happened.<\/p>\n","protected":false},"featured_media":9476,"template":"","meta":{"_acf_changed":false,"content-type":""},"resource-category":[77],"class_list":["post-9472","resource","type-resource","status-publish","has-post-thumbnail","hentry","resource-category-blog"],"acf":[],"_links":{"self":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource\/9472","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource"}],"about":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/types\/resource"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/media\/9476"}],"wp:attachment":[{"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/media?parent=9472"}],"wp:term":[{"taxonomy":"resource-category","embeddable":true,"href":"https:\/\/gapstars.net\/tech\/wp-json\/wp\/v2\/resource-category?post=9472"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}What went well:<\/h4>\n
\n
Where they struggled:<\/h4>\n
\n
StarPlanner \u2013 Team “FusionFocus”<\/strong><\/h3>\n
<\/p>\n<\/h2>\n
What We Learned<\/strong><\/h2>\n
Key Prompting Strategies That Worked<\/h4>\n
\n\u2192 Add login using email only
\n\u2192 Connect to Supabase project \u2018shuttle-prod\u2019 and fetch bookings<\/div>\n
\n\u2705 \u201cAdd filters for seat location (window\/aisle), floor level, and availability on hover\u201d
\n\u2705 \u201cUse Gapstars brand colors with a modern card-based layout\u201d<\/div>\ntest1<\/code> or demoTable<\/code> were used, the AI often defaulted to fake logic or broken code.<\/p>\n
\n\u2705 \u201cApply a minimal design with rounded corners, Gapstars Orange as primary CTA, and condensed spacing on mobile.\u201d<\/div>\n
\n\u201cRevert to V3 and reduce card padding\u201d<\/div>\nA Moment with Our Partners<\/strong><\/h2>\n
<\/p>\n<\/h2>\n
The Takeaway<\/strong><\/h2>\n
\n
<\/code><\/p>\n","protected":false},"excerpt":{"rendered":"