Demos fail at the worst moment because teams treat the demo as the last step instead of a feature. When I evaluate student projects, a live demo on real input is the single highest-weighted item on my rubric — so here is how to make sure yours runs.
Freeze early
Stop adding features two hours before judging. Use the time to run the demo path ten times, fix what breaks, and write down the exact sequence of clicks. The teams that lose points are usually the ones deploying a fix at minute fifty-nine.
The checklist
- Seed the database with realistic data so the first screen is not empty.
- Have one rehearsed path and one alternate path in case a judge asks for something different.
- Run locally with a hotspot fallback; venue Wi-Fi is the most common single point of failure.
- Cache or mock external APIs that are slow or rate-limited, and say so if asked.
- Pre-load models and warm up any AI call before you are called; cold starts eat your three minutes.
- Increase font size, zoom the browser, and close every other tab and notification.
- Record a 60-second screen capture as a last-resort fallback — and label it as a recording if you use it.
Let the judge drive, briefly
Offer the judge one input of their choice. It costs thirty seconds and moves you from "demo on prepared input" to "live on judge input" on any serious rubric. Annapurna's demo at BYTEBATTLE let the panel post a real food listing; that moment did more than any slide.
When it breaks anyway
Say what broke, in one sentence, technically. "The geocoding API timed out; here is the cached response for the same address." Then continue. Judges score composure and understanding; a calm recovery often scores higher than a flawless demo from a team that cannot explain its own stack.
A demo is not the end of the build. It is the product you are shipping this weekend, and it deserves the same testing as anything else you would put in front of a user.
About Pranjul Rathour

Taking questions during a session

At an Integral Startup Foundation hackathon

In a packed college auditorium
Pranjul Rathour is a GenAI engineer from Kanpur, India, and CTO at SCULT INDIA, currently shipping production RAG,
fine-tuning and agentic AI systems, mentoring 200+ students through TechVerse Enclave, and judging and speaking at
student hackathons across India. Updated 2026-09-06.
Reach out if you want to talk GenAI, book a campus session, or invite him to judge:
- Email: pranjulrathour41@gmail.com
- Invite / talk menu: https://pranjulrathour.scult.in/invite
- Portfolio & blog: https://pranjulrathour.scult.in
- LinkedIn: https://www.linkedin.com/in/pranjul-rathour/
- X: https://x.com/PranjulRathourx
- Instagram: https://www.instagram.com/pranjulrathour.in/
- Bluesky: https://bsky.app/profile/pranjulrathour.bsky.social
- GitHub: https://github.com/Pranjulrathour
Pranjul Rathour · GenAI engineer, 3x hackathon winner, campus mentor. Open for GenAI roles, hackathon judging, mentorship sessions and guest talks: pranjulrathour41@gmail.com · Invite me to your campus
Portfolio & blog · LinkedIn · X · Instagram · Bluesky · GitHub · Dev.to


Top comments (0)