Inspiration

A conference can be canceled. The need for community cannot.

The cancellation of Grace Hopper Celebration 2026 brought the motivation behind Diverse DMV into focus: career support should not depend on a single event taking place.

For people of underrepresented genders in tech, one conversation can open a door—a mentor offering guidance, a peer becoming a collaborator, or someone welcoming that first question. I wanted to create a place where those connections could happen throughout the year.

As a solo builder, I started with a focused community: Washington, D.C., Maryland, and Virginia. I built around one question: What are you here to do, and what are you open to helping others with?

What it does

Diverse DMV is a professional community for underrepresented genders in tech, helping members discover relevant people, connect with intention, and keep the conversation going.

Members can seek opportunities, offer mentorship, hire, or explore—and choose several intentions at once. A student looking for an internship can also help another beginner.

The application includes:

  • Purpose-driven profiles: Share skills, projects, what you need, and what you can offer. No résumé or LinkedIn account is required.
  • AI-powered discovery: Find members whose professional interests and offers relate to the connections you are seeking.
  • Mutual connections and messaging: Send an introduction, accept a request, and continue through persistent conversations.
  • Local office exploration: Explore corporate offices across the DMV through Google Maps, with regional and company-name search.
  • Visibility controls: Preview and publish your profile, manage audiences for sensitive fields, and remove or block connections.

The “Dare to Dive” design presents member profiles as boats in an ocean, accompanied by customizable fish. I wanted the first step into professional networking to feel approachable, with clear actions beneath the playful design.

How we built it

I built Diverse DMV solo as a full-stack application using Next.js, React, and TypeScript, with the interface and protected API routes in one codebase.

Firebase Authentication handles account creation and login, supported by server-verified, revocable sessions. Tiger Data’s PostgreSQL database stores profiles, requests, connections, and message history.

For discovery, Google Gemini embeddings convert permitted professional profile text and connection goals into 768-dimensional vectors. pgvector retrieves relevant members using cosine similarity, while server-side rules exclude blocked, unpublished, and otherwise ineligible profiles.

Google Maps JavaScript API and Places API power the office explorer. Custom SVG illustrations, CSS animations, and responsive components create the ocean-inspired interface.

I verified the core journey through automated tests and browser walkthroughs with real providers: publish a profile, discover someone, connect, exchange messages, and return to the conversation after logging in again.

Challenges we ran into

Building solo meant connecting product design, frontend development, backend logic, and testing into one coherent experience. Keeping the scope centered on a complete connection journey helped me make those decisions.

The hardest technical challenge was keeping discovery aligned with privacy. Recommendations must respond when someone changes visibility, unpublishes a profile, or blocks another member. I enforced those boundaries on the server and versioned embeddings to prevent outdated profile data from remaining eligible.

Messaging brought another subtle challenge: a message can save successfully even when its confirmation never reaches the browser. I used stable operation identifiers and database transactions so retrying would not create duplicates.

I also had to balance the animated ocean design with practical usability—readable contrast, visible keyboard focus, mobile layouts, and reduced-motion support.

Accomplishments that we're proud of

I’m proud of taking Diverse DMV beyond a visual prototype into a working application connecting real authentication, database, AI, and mapping services.

The core experience works end to end: members can publish profiles, receive recommendations, form mutual connections, and exchange messages that persist across sessions. I also tested important boundaries, including denying conversation access to unrelated accounts and preventing duplicate messages during retries.

I’m equally proud of making room for beginners. Someone without a polished résumé or established professional network can still express what they want to learn and what they can contribute.

What we learned

I learned that useful professional discovery begins with better questions. Credentials describe someone’s past; intentions help reveal which conversation could matter next.

I learned that privacy is an architectural decision. It affects what enters an embedding, which profiles appear in recommendations, and what remains accessible when a relationship changes.

Building solo also taught me to prioritize complete, testable experiences. Connecting two people reliably required attention to everything between their first visit and their next conversation—including expired sessions, interrupted requests, and unavailable services.

What's next for Diverse DMV

My next milestone is a hosted pilot with local students, professionals, and community groups. Before a broader launch, I plan to complete deployment verification, production email recovery, remaining accessibility checks, and production Maps configuration.

After that, I want community feedback to guide richer project showcases, clearer mentoring availability, and user-reviewed résumé import. Establishing a practical moderation process will also be essential as the community grows.

I started Diverse DMV to make the next meaningful career conversation easier to begin—and give it somewhere to continue.

Built With

Share this project:

Updates

Submission history