FinHealth

Home Dashboard
AI Insights & Latest News
Chatbot
Tailored Investment and Analysis Form
Tailored Investment and Analysis Results
Portfolio Analysis

Inspiration

Managing your finances can be pretty intimidating, especially if you’re just starting out. We were inspired to create a platform that makes it easier to understand investments and stay on top of your financial goals. That’s how FinHealth came to life—a user-friendly tool that aims to help you feel more confident about your money decisions.

What it does

FinHealth acts like your personal financial assistant, all in one place:

Stock Recommendations: Our dashboard highlights recommended stocks and shows related news with sentiment analysis (positive, neutral, or negative).
Context-Aware AI Chatbot: This chatbot doesn’t just answer random questions—it tailors its responses to the stocks you’re tracking, recent market news, or your personal questions.
Tailored Investment Plans: By entering your personal financial situation and investment goals, FinHealth creates a customized plan that aligns with your vision and risk tolerance.
Portfolio Analysis: Even if you only have a screenshot of your portfolio, FinHealth can parse it to provide deeper insights and actionable advice.

How we built it

Tech Stack & Framework: Built primarily with Streamlit (Python). We also leveraged libraries such as OpenAI, pandas, yfinance, scikit-learn, nltk, and plotly to power various features.
Real-Time Data: We rely on Yahoo Finance for up-to-date stock quotes and historical data.
AI & NLP: OpenAI's GPT-4o-mini model powers our chatbot’s contextual conversations, while a Sentiment Analyzer interprets recent market news. As for generating buy/hold/sell suggestions, we used a Random Tree Classifier.
UI Enhancements: We used custom HTML/CSS within Streamlit components to achieve a clean, modern look and feel.

Challenges we ran into

Learning Curve with Streamlit: Since it was our first time using Streamlit as a framework, setting up a fully interactive, real-time frontend came with its own set of challenges.
Achieving Satisfactory Accuracy: Building a reliable stock prediction model is no small feat. We managed to hit an average of 79% accuracy within 24 hours, thanks to multiple iterations and fine-tuning.

Accomplishments that we're proud of

Integrated, Easy-to-Use Dashboard: Bringing real-time stock data, AI-driven insights, and a contextual chatbot together in one interface.
Tailored Investment Analysis: Allowing users to submit their financial details for a personalized plan that aligns with individual risk profiles.

What we learned

Effective AI Integration: Crafting the right prompts and logic flow for the chatbot is crucial to delivering relevant, reliable advice.
Balancing Features and Simplicity: Combining sentiment analysis, news aggregation, and an AI chat in one platform taught us a lot about delivering a polished user experience.

What's next for FinHealth

Deeper Analytics: Adding more metrics and visualizations to give users better portfolio insights.
Mobile Compatibility: Optimizing for smartphones and tablets so people can manage their finances on the go.
User Accounts: Implementing user profiles so portfolios and preferences can be saved and tracked over time.