TIEHub Origin Weekend (Hackathon)

Introduction

Imagine building a startup in under a week. At USC’s Technology Innovation and Entrepreneurship (TIE Hub) Origin Weekend, that’s exactly what we were challenged to do. This 5-day startup sprint was hosted by the Viterbi School of Engineering and the Lloyd Greif Center for Entrepreneurial Studies. The program was designed to take us from idea to venture through workshops, mentorship, and a final pitch competition for $5,000 in prizes. More than 375 students from across USC joined in and somehow, my team made it to the final round.

Day 1: Hackathon Kickoff & Team Formation

Thursday evening kicked off with an orientation and problem prompts. We heard from Jennifer Roberts of Grit VC and Alex Lee of Truewind (YC S25), both of whom drove home a lesson that carried through the entire event: great startups don’t start with grand ideas, they start with painful, real problems worth solving.

Alex Lee’s advice really stuck with me. He said most customers can only name three pain points they actually care about. If your idea addresses pain point number 9 or 10, it probably won’t matter to the customer. That simple reminder helped me look at problems with more focus: Who exactly are we helping, and do they care enough about this issue for us to matter?

The night ended with team formation activities. It was chaotic but energizing. After bouncing between conversations, I joined a team made up of two computer science majors, one electrical engineering major, one astronautical engineering major, and myself, a mechanical engineering major. Our multidisciplinary group focused on tackling the problem of extreme environmental impact. That’s where our project, GreenCast, took shape as an AI-powered platform that visualizes and gives quantifiable data as to how green spaces reduce urban heat and other climate risks. The idea felt both urgent and practical, with clear customers and identified pain points that I’ll cover in more detail below.

Day 2: Validation, MVPs, and Business Models

Friday’s first workshop was one of the most eye-opening experiences for me, because we were introduced to the new generation of AI-driven tools that allow startup founders to quickly bring a minimum viable product (MVP) to life.

The workshop sessions introduced us to a layered framework of tools:

• Foundation model layer: ChatGPT, Anthropic, Deepgram — the brains for text and voice.

• No/low-code automation: n8n, Zapier, Langflow — the interface that glues everything together.

• No-code web & app builders: Bubble.io, Replit, V-Zero — front-end and dashboards.

• AI-first IDEs: Cursor, Windsurf, Claude Code — speeding up actual coding when needed.

To demonstrate how we can apply all of these tools, the instructor, ​Raj Vasantha (Co-Founder @ SynthioLabs & YC X25), walked us through a voice reservation assistant for restaurants.

1. User calls a phone number, handled by a Vapi.AI voice bot.

2. Conversation is transcribed and parsed in n8n, which converts natural speech into structured JSON data (like time: 7:30 pm, number of guests: 4).

3. Airtable stores the reservation as a record in a database.

4. Bubble.io provides a dashboard so staff can view reservations and send confirmations.

What impressed me was not just the cleverness of the stack, but how accessible it was. No specialized coding required, just plugging tools together like Lego blocks.

For me, it was the first time I truly saw how automation could transform my productivity as an engineer. Why spend hours on repetitive tasks when I could design AI-driven assistants to handle them? This experience opened the door to a future project I am very excited to share! Stay tuned for a blog post or new project where I show off what I was able to achieve with the tools given above.

Day 3 & 4: Building the Idea. From Theory to Execution

Saturday and Sunday were the grind days. At first, our team struggled to shape an MVP around the original space prompt. Tackling food production systems for astronauts was ambitious, but it quickly became clear that it wasn’t a problem we could solve in just two days, and was not necessarily a top 3 pain point for space travel at the moment. After some tough conversations, we decided to pivot toward a problem that felt both urgent and actionable here on Earth: extreme environmental impact.

That’s how GreenCast.ai took shape. Our idea was to build an AI-powered platform that quantifies and visualizes how green spaces reduce urban heat and other climate-driven risks. The next challenge was execution. What would our MVP look like by Sunday night? We debated whether to prioritize visuals, data, or storytelling, but ultimately decided that clarity mattered most. Using the six-slide pitch structure provided by the organizers, we distilled our work into:

1. The urban heat problem.

2. Customer insights from city planners, nonprofits, and environmental groups.

3. A web-based version of GreenCast.ai as the MVP concept.

4. Why it was different: simple, affordable, and actionable compared to complex existing tools.

5. A business model centered on a freemium approach with subscription-based pricing, designed to serve B2B, B2C, and B2G markets.

6. Next steps for testing and validation.

By the time we submitted our deck on Sunday night, it finally felt like we had something real. The pivot wasn’t easy, but it gave us a solution that felt urgent, practical, and grounded in reality.

Click the image above to access the MVP website OR Click the following hyperlink to view a live demo.

Day 5: Final Pitch Night

Out of more than 50 teams, only 15 were selected as finalists to pitch live in front of an audience of peers, mentors, and a panel of VCs and founders. Our team made the cut, which meant we had just 3 minutes to present GreenCast.ai and 2 minutes for Q&A.

Each team had a different take on the prompts, ranging from social platforms to climate tech to AI-driven tools. When it was our turn, we walked on stage with a mix of nerves and excitement. We laid out the urban heat problem, our customer insights, and GreenCast.ai’s ability to quantify and visualize the cooling power of green spaces. We closed by sharing our freemium subscription model and next steps for validation.

While we didn’t win the $5,000 prize, being a finalist and presenting live was an achievement in itself. It gave us a chance to showcase our work in front of real investors and gain feedback from professionals who have been in our shoes before. More than anything, it felt like proof that in just five days, we had gone from strangers with no idea to a team with a tangible product and a vision worth pitching.

Reflection & Lessons Learned

Looking back, Origin Weekend was more than just a startup competition. It was a crash course in teamwork, resilience, and problem-solving under pressure. A few key lessons stood out:

• Start with pain points, not ideas. The most valuable advice from mentors and speakers was to anchor everything in customer pain.

• MVPs are about clarity, not perfection. We had to resist the urge to overbuild. The winning teams weren’t the ones with the flashiest tech, but the ones with the clearest solutions and business models.

• Cross-disciplinary teams are powerful. Having computer science, engineering, and design perspectives in one group made our approach stronger. Everyone brought something different to the table. We also became great friends by the end of the hackathon and now have a variety of skills to leverage in any future ventures we take on.

• Presentation is half the battle. Compressing days of work into a three-minute pitch forced us to sharpen our storytelling and highlight what mattered most. The most important lesson of all:

Showmanship is your #1 asset in selling your idea to investors.

Even though our team did not win, Origin Weekend left me with new skills, new collaborators, and the confidence to take on future projects. Most importantly, it sparked ideas for how I can use AI and automation not just in competitions, but to make a meaningful impact in my own engineering work and daily life.