Inspiration
As an introvert, I’ve always felt that networking events are full of opportunity, but not always easy to navigate. Many people, especially students, fresh graduates, and SMB owners in Malaysia, walk into career fairs or business events without knowing how to introduce themselves, who to approach, or how to explain their value clearly.
That inspired us to build AI Networking Copilot. We wanted to create something that does more than summarize profiles. We wanted an AI tool that helps people understand their own strengths, read the room better, and start more meaningful conversations with confidence.
What it does
AI Networking Copilot is a Gemini-powered mobile-first app that helps users prepare for career fairs and networking events.
It can:
- analyze the user’s own profile and generate Spark Points, Exchange Value, and a short self-introduction
- understand event context from an event type, event name, or uploaded artifacts like an event banner or schedule image
- identify and analyze the person the user wants to meet through name search, pasted profile text, LinkedIn-style info, or multimodal inputs like a name card, resume screenshot, or booth poster
- generate Mutual Relevance, natural conversation starters, follow-up questions, and suggested next steps
The goal is to turn networking from a stressful guessing game into a more confident and practical experience.
How we built it
We built AI Networking Copilot as a mobile-first PWA using:
- React
- Vite
- Tailwind CSS
- LocalStorage for lightweight local persistence
We designed the app around three main flows:
- Know Me – understand the user’s own strengths and value
- Know the Event – understand the context of the room
- Bridge Us – connect the user with the person they want to meet using Gemini reasoning
Gemini was used not just for basic text generation, but for combining multiple sources of context:
- user profile information
- target person information
- event context
- multimodal inputs such as name cards, resumes, booth posters, and schedule images
This allowed us to generate more relevant, event-aware, and practical networking suggestions.
Challenges we ran into
One of the biggest challenges was making sure the project felt like a real AI solution instead of just a text-in, text-out wrapper. We had to push beyond simple summarization and focus on reasoning, mutual relevance, and event-aware conversation strategy.
Another challenge was scope. We had many possible input types and user flows, so we had to stay disciplined and focus on a hackathon-sized MVP.
We also had to think carefully about the real user experience. In a live event setting, people need something fast, simple, and useful on mobile. That forced us to keep the interface lightweight and practical.
Accomplishments that we're proud of
We’re proud that we found a way to turn a very human problem into a practical AI product.
A few things we’re especially proud of:
- grounding the project in a real Malaysian problem instead of building a generic productivity tool
- designing the app around underserved users like students and SMB owners
- using Gemini in a more meaningful way through multimodal input, longer context, and reasoning across multiple signals
- creating a product concept that feels personal, useful, and easy to demonstrate
We’re also proud that the idea came from a real emotional pain point: the feeling of having value but struggling to express it in the room.
What we learned
We learned that the hard part of networking is not access to information. It is turning messy information into useful human judgment.
We also learned that AI becomes much more valuable when it helps with:
- confidence
- positioning
- relevance
- social strategy
Another big lesson was that local relevance matters. The project became much stronger once we focused on Malaysian students and SMB owners, instead of trying to build a generic global networking tool.
Technically, we also learned the importance of balancing ambition with clarity. A simple, well-defined product flow is much stronger than too many disconnected AI features.
What's next for AI Networking Copilot
Next, we want to expand AI Networking Copilot in a few directions:
- add stronger multilingual support for English, Bahasa Malaysia, and Chinese
- improve person matching and candidate selection
- add a post-meeting follow-up mode for drafting messages and tracking connections
- support richer offline storage with IndexedDB
- explore organizer-side matchmaking for career fairs, networking events, and SMB expos
Longer term, we see AI Networking Copilot becoming a broader opportunity-access tool for Malaysia: helping more people show up prepared, communicate their value clearly, and turn conversations into real career or business outcomes.
Log in or sign up for Devpost to join the conversation.