Paloor

Your AI Finance Pal - Learn to Invest with Confidence

Comment

Paloor: Democratizing Financial Education Through AI

Inspiration

My father grew up in a distant village in southern India called Kizhapalaiyur (கீழப்புலியூர்), Palur for short. His father (my grandfather) never attended schooling beyond middle school. And yet, through sheer determination and an unshakeable belief in education, he managed to send all eight of his children, four sons and four daughters, not just to college, but to graduate programs and professional certifications as well. No financial advisor. No inherited wealth. Just a vision.

That vision is why I named this platform Paloor, to carry his legacy forward.

I studied Finance and Economics at Emory University, graduating summa cum laude, worked at a private equity portfolio company with $300M+ ARR, then at a mid-market investment bank. I'm currently completing a Master's in Applied Data Science and interning at CBRE Investment Management as a Data Scientist. By every conventional measure, I am exactly the kind of person the financial system was built for.

And that's precisely the problem.

Finance is not a zero-sum game. When more people understand how to manage money and navigate markets, everyone benefits. But the tools and mental models were designed by insiders, for insiders. When I started talking about Paloor, business school graduates who aced college-level finance told me they were still scared to open a brokerage account. Friends who graduated with me from business school said they avoided reading their 401 (k) statements. The fear wasn't about intelligence. It was about access. Access to someone patient enough to explain it plainly, without judgment, at any hour.

That is the gap Paloor exists to close. Not a financial product. A financial educator.

A Note on Honesty

This is my first hackathon. I am not a software engineer by training, so I taught myself how to and set up an AWS infrastructure and backend foundation prior to the event beginning to get myself ready. The Alpha Vantage API key and core AWS backend were in place at the start of the hackathon.

Everything I am submitting for judging: the Learning module with ElevenLabs voice, the Gemma-powered quiz loop, the contextual Paloor AI on the equities desk, the screenshot-to-chat capability, and the Deep Analysis agentic pipeline. This was all built by me during the hackathon.

What I Built

1. The Learning Module: Voice-First Financial Education

The learning module is built around a fully voice-driven loop:

  • Google Gemma acts as the curriculum engine, generating lessons (vetted by me) from first principles, adapting as the student progresses, and writing checkpoint questions that test genuine understanding.
  • ElevenLabs converts every lesson and every piece of feedback into natural, expressive speech.

The key design decision was full voice bidirectionality. A student listens to the lesson, answers the quiz out loud, and receives spoken feedback, one uninterrupted flow, no typing required. For a student with a reading disability, a language barrier, or who simply learns by listening, this is the difference between a platform that works for them and one that doesn't.

2. The Equities Desk: Learning on Real Data

Rather than invented textbook examples, Paloor teaches with live market data via Alpha Vantage. Paloor AI, powered by Gemma, sits alongside the data with full context. It already knows which ticker you're viewing and answers questions conversationally, like a teacher sitting next to you.

The feature I'm most proud of: screenshot-to-chat. Drag a selection over any chart on the page, drop it into the chat, and Gemma explains it in plain language. You don't need to know what a Bollinger Band is called to ask about it, you just click it. That's how intuition gets built.

3. Deep Analysis: A Masterclass in Professional Reasoning

An agentic pipeline modeled after a real trading desk. When triggered, it spawns Gemma-powered agents running in parallel:

Agent Focus
Market Analyst Price action, trend, 52-week context
Technical Analyst RSI, MACD, Bollinger Bands, moving averages
Fundamentals Analyst P/E, revenue growth, balance sheet
News Analyst Recent sentiment and catalysts

Their reports flow through a Research Debate, a Trader, a Risk Manager, and a Portfolio Manager who issues a final verdict. The final synthesis can be expressed as:

$$V = f(s_{\text{market}},\ s_{\text{technical}},\ s_{\text{fundamentals}},\ s_{\text{news}})$$

For a student, the output isn't just an answer — it's a transparent window into how professionals think. Seeing that process demystifies markets. That is the educational outcome.

Tech Stack

Layer Technology
LLM Google Gemma
Voice (TTS) ElevenLabs
Voice (STT) Amazon Transcribe
Market Data Alpha Vantage API
Backend FastAPI, PostgreSQL, AWS (S3, Secrets Manager)
Frontend Next.js, React, Tailwind CSS, shadcn/ui

Challenges

The hardest thing I built was the voice loop, getting ElevenLabs, Transcribe, and Gemma to form a seamless, low-latency round-trip required careful async handling and a lot of iteration.

The most humbling thing was how much I didn't know about building software. I've spent years in finance building models and decks. Debugging a FastAPI dependency injection issue at 4 am is a completely different skill set. I learned more about building this weekend than in months of self-study.

What's Next

Paloor is a beginning. The vision is a platform where anyone, regardless of whether they grew up in a financially literate household, regardless of language or ability, can genuinely learn how to think about money. Not be told what to do with it. Learn to think about it.

My late grandfather from Kizhapalaiyur believed that education was worth everything. My dad believed him, and I believed him. This platform is his legacy, just built with the tools of today.

Built With

Share this project:

Updates