Maven-Focus on Empowerment

Inspiration

We realized that while talent is distributed equally, opportunity and guidance are not. "Maven" means an expert or connoisseur, and we were inspired by the idea that everyone deserves access to a Maven to help them navigate their career, education, and personal growth. We wanted to build a platform that focuses entirely on empowerment—giving users the tools, guidance, and community they need to upskill and confidently take up space in their respective industries.

What it does

Maven is an AI-powered empowerment ecosystem designed to bridge the gap between ambition and actionable success.

• AI-Powered Roadmaps: Users input their current skills and ultimate goals, and our AI generates a step-by-step, personalized learning roadmap.
• Smart Mentorship Matching: We connect mentees with "Mavens" (industry experts) using a sophisticated similarity algorithm to ensure high-quality, relevant mentorship.
• Empowerment Hub: A resource pool for interview prep, confidence building, and financial literacy, accessible even in low-resource environments.

The Math Behind the Matching

To ensure mentees are paired with the most empowering mentors, we developed a weighted scoring algorithm. The match efficiency is calculated using a similarity index between a user's target goals and a Maven's demonstrated expertise.

Let $S_{user}$ represent the user's desired skill vector and $S_{maven}$ represent the mentor's expertise vector. The Empowerment Match Score ($M$) is calculated as:

$$M = \alpha \cdot \frac{S_{user} \cdot S_{maven}}{|S_{user}| |S_{maven}|} + \beta \cdot A(t) - \gamma \cdot D$$

Where:

• $\alpha, \beta, \gamma$ are tunable weight parameters.
• The first term calculates the cosine similarity of skills.
• $A(t)$ is the time-availability matrix of the mentor.
• $D$ represents the experience gap delta (ensuring the mentor isn't too far removed from the mentee's current stage).

How we built it

We utilized a modern, scalable tech stack:

• Frontend: Built with React.js and Tailwind CSS to ensure a responsive, highly accessible, and user-friendly interface.
• Backend: Powered by Node.js and Express.js for robust API routing.
• Database: MongoDB for flexible, NoSQL data storage of user profiles, roadmaps, and chat logs.
• AI Layer: We integrated the OpenAI API via Python to dynamically generate the personalized growth roadmaps and process natural language inputs from users.

Challenges we ran into

One of the core challenges was fine-tuning the AI prompts to ensure the generated roadmaps were actionable rather than generic. Additionally, integrating the Python AI microservice with our Node.js backend required us to implement a seamless message broker system to prevent API timeouts during heavy query loads.

What we learned

We learned a tremendous amount about AI prompt engineering, specifically how to constrain LLMs to output structured JSON data for our frontend to render beautifully. We also gained deep insights into accessibility standards (WCAG), ensuring our empowerment platform is truly usable by everyone.

What's next for Maven

We plan to introduce Maven Mobile (built on React Native) to reach users who only have access to smartphones. We also want to implement peer-to-peer "micro-mentorships" and integrate a verifiable credential system using blockchain to certify the skills users gain on the platform.

Built With

• express.js
• javascript
• mongodb
• node.js
• openai
• python
• react.js
• tailwind-css