Nebula

AI-powered CRM for personal networking that remembers your connections, every interaction, and tells you exactly when and how to follow up.

Comment

About the Project

As MBA students, we were inspired by how challenging it is for busy professionals to stay on top of networking: remembering conversations, follow-ups, personal details, and next steps. Networking is one of the most important parts of career growth, yet it often becomes fragmented across emails, notebooks, LinkedIn messages, and memory.

During the hackathon, we explored how AI workflows can be used to reduce that friction. Our idea was simple:

What if your network could stay warm without you having to manage it?

What We Built

We built our functional prototype using Lovable, focusing on:

  • A clean, intuitive user flow that feels natural and minimal.
  • A real backend and database that securely stores and updates all contact data.
  • A voice AI that records interactions, transcribes them, and generates summaries, insights, and follow-up actions automatically.
  • An AI chatbot that understands networking needs, searches contacts, recommends who to reach out to, and drafts messages or next steps in natural language.
  • A rich, searchable contact system with customizable tags, interests, and deep interpersonal context. (including how you can support someone, how they can help you, shared history, and personal details that matter)

The product experience follows a simple philosophy:

Record an interaction → the system handles everything else.

What We Learned

We learned how important scoping is when building with AI. Our biggest challenge was narrowing the vision into something that could be built meaningfully in 48 hours. We iterated quickly, prioritized core value, defined the MVP, and leaned on autogenerated workflows to accelerate development.

We also learned how much people resonated with the problem. Nearly everyone we spoke to immediately responded with:

“I need this.”

Challenges Faced

  • Balancing a big vision with limited time
    We had far more ideas than hours, which meant ruthless prioritization. Choosing what not to build became just as important as defining what belonged in the MVP.

  • Designing a system that feels personal, not robotic
    Because this product supports human relationships, tone and experience mattered. We iterated on UX, messaging, and interaction design to ensure the interface felt helpful, warm, and human.

  • Building data structures that can scale with evolving user context
    We understand that contacts aren’t static records: roles change, interests evolve, relationships deepen. Structuring tags, relationship metadata, and longitudinal history in a flexible way required thoughtful database design.

  • Making AI output feel trustworthy, accurate, and actionable
    We needed the AI to produce reliable summaries, smart recommendations, and meaningful next steps. Quality mattered more than quantity, and we realized that how we prompt the system directly impacted output clarity, tone, and usefulness. Designing and refining prompts became as critical as the model itself.

  • Keeping complexity invisible to the user
    With many features (voice logging, summaries, reminders, tags, search, and a chatbot), the challenge was maintaining simplicity. The experience needed to feel effortless, even though much intelligence was happening behind the scenes.

Why This Matters

Professional relationships shape careers, but remembering details, follow-ups, and context shouldn’t require constant cognitive load. We want networking to feel natural, humane, and effortless.

Our prototype is a step toward:

  • Automatic memory
  • Automatic follow-through
  • Automatic relationship context

Or in shorter terms:

A new brain for professional relationships.

Final Reflection

We’re proud of what we built. We stretched our thinking, collaborated deeply with each other, learned new tools, and shipped something real under time pressure.

Most importantly we had fun! 🚀

Instruction for trying out the prototype

Feel free to watch our demo video on YouTube for a quick walkthrough of how to navigate Nebula. When prompted to log in, please use the sample credentials below:

This step is required because the backend is fully functional, and new account creation is disabled for demo purposes. Logging in with the sample account will allow you to explore the existing data and try all key features.

Built With

  • lovable
  • openai
  • supabase
Share this project:

Updates