Mentor Bridge

Mentorship reimagined for women in tech. AI that matches you with mentors who get your goals, your voice, and your journey.

Comment

Inspiration

We’ve all had moments when we wished someone could just get it. Whether it’s navigating a new career, breaking into a male dominated field, or simply trying to find your voice, mentorship can make all the difference. But most platforms feel too generic, like cold directories, not communities.

Mentor Bridge was born from a need for something more intentional, a safe, empowering space for women to connect with mentors who truly understand them. We set out to go beyond matching and instead build a relationship focused platform with thoughtful tools, emotional check-ins, and a supportive community, backed by smart AI where it actually adds value.

What it does

Mentor Bridge is a mentorship platform built for women across all career stages and fields, designed with role-aware personalization, AI powered features, and a gentle, community-driven UX. Here's what it offers:

👥 Role-Based Sign-Up: Mentors and mentees create rich, career-aware profiles with experience levels, tags, preferences, and mentoring goals, all stored in a flexible MongoDB schema for dynamic filtering and matching. This ensures tailored connections that respect where users are in their journey.

🔍 Browse or Match: Users can explore profiles using filters or opt for AI-powered matches, which suggest mentors based on personality, mentoring style, and goals. Matching is powered by Gemini and gives thoughtful reasons behind each suggestion, making discovery more intentional and less overwhelming.

🔐 Secure Auth with Auth0: We implemented role-specific authentication with Safe Mode, allowing users to hide sensitive info like contact details or photos. This helps users feel safe and in control, especially when navigating vulnerable conversations or early-career stages.

💬 In-App Messaging + Gemini Intro Assistant: To reduce the anxiety of cold outreach, mentees can use Gemini to help draft confident, friendly first messages. The experience is integrated with StreamChat for seamless, real-time conversations that don’t require storing messages on our end.

🌸 Anonymous Ask Wall: A supportive community feature where mentees can anonymously ask questions and receive encouragement or advice. Questions are public, reactions are built-in, and no identifiable info is stored, creating a safe space for candid conversation.

🧠 Conversation Coach: When mentees need to ask tough questions (like about salary, referrals, or career pivots), the Conversation Coach uses Gemini to help reframe them with clarity and tact, turning hesitation into healthy communication.

✨ Affirmation Mode: Mentorship isn’t just about career growth, it’s emotional too. This mode offers daily check-ins and uplifting AI generated affirmations based on how the mentee is feeling, supporting mental wellness as part of the experience.

🎉 Celebrate Your Win: Big or small, progress matters. Users can log milestones (like landing an interview or finishing a project) and receive personalized encouragement, a small feature that goes a long way in building confidence and reflection habits.

🧩 StreamChat Messaging: In-app messaging lets mentors and mentees chat seamlessly, making conversations feel personal, real-time, and easy to keep going without needing to store the messages.

How Matching Works

Mentor Bridge uses a two-phase matching system designed to balance structure and emotional intelligence:

Phase 1: Initial Compatibility Scoring (Used in Browse Mode & Match Mode) We assign a base score out of 100 using a weighted formula:

  • Language, industry, domain, and tech stack overlaps: up to 15 points each
  • Mentee type alignment: 10 points
  • Availability: 10 points
  • Cosine similarity between mentor/mentee profile vectors: up to 20 points

This score helps rank mentors even before AI is involved, offering mentees personalized discovery right from Browse Mode.

Phase 2: Gemini Refinement (Used in Match Mode) In Match Mode, we send both profiles and the initial score to Gemini (Google Generative AI), which evaluates deeper alignment like:

  • Shared goals and motivations
  • Compatibility of communication and mentoring styles
  • Relevant experience that supports the mentee’s needs

Gemini responds with a final match score and 2–3 personalized reasons for why the pair would work well, bringing emotional nuance into the algorithm.

How we built it

  • Frontend: Flutter (fully responsive UI for mobile + web), StreamChat for real-time messaging
  • Backend: FastAPI (Python 3.11), MongoDB (Atlas), AWS Services (API Gateway for routing, IAM roles for secure access, and CloudWatch for monitoring), Boto3, StreamChat for messaging. Deployed on AWS Lambda using the Serverless Framework.
  • Authentication: Auth0
  • AI features: Integrated the Google Gemini API to enable AI-assisted features:
    • Intro Assistant for crafting confident opening messages
    • Conversation Coach to help mentees reframe difficult questions
    • Affirmation Mode for daily emotional check-ins and encouragement
  • Matching Logic:
    • A custom scoring algorithm based on profile tags, availability, and vector similarity
    • Gemini refinement for context-aware scoring and tailored reasons for each match
  • Storage: MongoDB (for user data, asks, and chats), AWS S3 (for optional profile photos)

Key highlights:

  • Human-first matching that blends smart scoring with AI powered insights to create thoughtful, personalized connections.
  • Affirmation Mode for emotional wellness, daily check-ins and AI-generated encouragement designed to uplift.
  • Celebrate Your Win feature that helps mentees log milestones and receive encouraging feedback, reinforcing progress.
  • Inclusive design that supports women across all career stages, industries, and identities, from first-gen students to senior leaders.
  • Anonymous Ask Wall to foster peer support and honest questions in a safe, non-judgmental space.
  • Built with care, using ethical AI and privacy-conscious architecture that centers trust and empowerment.

Challenges we ran into

  • This was our first time using Auth0, MongoDB Atlas, and Google Gemini. Understanding role-based authentication, modeling flexible user data, and structuring AI prompts took experimentation and iteration.
  • We had to carefully define which endpoints were public or protected, especially with sensitive data like user profiles and anonymous questions, a foundational but critical decision.
  • Creating clean, separate experiences for mentors and mentees in both frontend and backend required writing custom Auth0 rules and tailoring token-based logic.
  • Making it safe yet engaging was tricky, we didn’t want it to feel like an empty feed or a black hole. Structuring reactions, moderation logic, and visibility helped.
  • Scoping with empathy: We brainstormed many powerful ideas (session booking, async voice notes, calendar sync), but had to prioritize impact and feasibility to ship a meaningful MVP on time.

Accomplishments that we're proud of

  • Creating a role-aware matching engine that blends logic and emotional nuance, making mentorship feel like a connection, not just a calculated pair.
  • Bringing AI into the experience thoughtfully, with features like the Intro Message Assistant and Affirmation Mode that genuinely support users in human, non-cringey ways.
  • Designing for safety and trust, from the Anonymous Ask Wall to customizable profile visibility, building a space where mentees feel seen and secure.
  • Building a fully serverless architecture on AWS, integrating FastAPI, MongoDB, Auth0, Gemini, and StreamChat into one cohesive system that runs smoothly and scales cleanly.
  • And most of all, building a space by women, for women, that’s intentional, intelligent, and deeply human.

What we learned

  • AI + empathy = magic: AI works best when it augments human connections, not replaces them.
  • Designing for safety and identity at the same time is complex, but so worth it.
  • Serverless FastAPI + MongoDB + Gemini is a surprisingly powerful stack for agile social apps.
  • Good UX isn’t just buttons and fonts, it’s emotional tone, community cues, and thoughtful defaults.

What's next for Mentor Bridge

We’re just getting started. Here's what’s next:

  • 📅 Session Scheduling: Integrated with calendars and time zones.
  • 🔔 Push Notifications: For messages, reminders, and affirmations.
  • 🗂 Mentorship Reflection Journal: A space for mentees to log insights and growth.
  • 🪞 AI Mirror: A feedback bot that reflects on your mentorship journey and growth areas.
  • 🏘 Group Circles: Small themed pods like “Women in Tech in College” or “Working Moms in STEM”.

Built With

Share this project:

Updates