We noticed a huge gap in how financial literacy is taught — it's often dull, generic, and disconnected from the real financial decisions people make every day. Most tools feel like textbooks, not companions. We wanted to change that.
We envisioned FinLearn AI as a daily, personalized financial learning hub — one that makes understanding finance engaging, relevant, and actionable. We were inspired by how people already consume bite-sized content (like social media reels or news notifications) and thought: why can’t learning money be just as intuitive and habit-forming?
FinLearn AI is a personalized financial literacy companion that empowers users to:
📚 Learn daily from AI-curated news articles, simplified into actionable lessons.
🧠 Understand tough financial terms via smart tooltips.
📈 Track a personalized watchlist for crypto and stocks.
🔥 Build a learning habit with streaks and quizzes.
Frontend
- Next.js + TypeScript: For a fast, secure, and scalable web app.
- Tailwind CSS: To design a responsive and clean UI quickly.
- Framer Motion: For fluid animations that enhance user engagement.
- Redux: To manage global state (preferences, watchlist, streaks).
Backend & Infrastructure
- Python + FastAPI: To build a high-performance backend for APIs and business logic.
Firebase
- Auth for secure login and user identity
- Firestore to store user progress, preferences, and goals
AI & Content Layer
- Perplexity AI API:
- Fetches and simplifies news articles
- Generates customized financial tips and learning modules
- Powers personalized story-based lessons using real-world data
🧠 Crafting Truly Personalized Learning
Generating AI-powered content that felt genuinely personalized to each user’s goals and preferences required thoughtful formatting and adaptation.
📚 Explaining Financial Jargon in Context
Implementing tooltips for dynamically highlighted financial terms within AI-generated content demanded a robust, non-intrusive parsing system.
🎯 Turning Passive Learning into Action
Bridging the gap between learning and goal-setting involved designing intuitive flows that motivated users to immediately act on new financial knowledge.
🔄 Real-Time Data Integration
Syncing real-time crypto and stock data with user watchlists while maintaining performance required optimized fetching and smart caching strategies.
- Built a full-stack AI-driven learning platform that makes finance practical and engaging
- Integrated Perplexity SONAR APIs for real-time content transformation from finance news
- Created a fully responsive, animated UI with Framer Motion + Tailwind
- Developed intelligent goal-tracking and progress feedback systems
- Made finance more approachable for users who previously found it intimidating or boring
🧩 Gaps in Existing Financial Apps
We learned that most financial apps fail to connect education with real-world action, leaving users unmotivated and underserved.
🤖 Working with Perplexity AI API
We discovered that while Perplexity SONAR APIs provides rich content, its limitations around structure and latency required us to build fallback systems and formatting layers to maintain quality.
📈 Personalization is More Than Preferences
We realized that meaningful personalization must evolve with user behavior — like article interactions, quiz results, and goal completions — not rely solely on static inputs.
⚙️ Micro-Interactions Drive Macro Habits
Small design elements like tooltips, streaks, and nudges significantly boosted engagement and helped users build consistent financial habits.
📱 Launch a mobile app using React Native to extend accessibility
🧑🤝🧑 Add community challenges where users can learn and grow together
💬 Introduce a chat-based AI assistant for finance-related queries in real time
🔐 Build goal-driven learning tracks like “Debt-Free Journey” or “First-Time Investor”
🏦 Integrate with budgeting tools and allow real-time sync with bank/UPI data
- Bhavya Jain - LinkedIN
- Athul - LinkedIN
- Rohan Singla - @rohanBuilds
FinLearn AI was developed as part of the Perplexity AI Hackathon.
This project is licensed under the MIT License.