AI-stock-screener

Screen-Finance — Smarter Stock Research, Powered by Agentic AI. Analyze Indian & US stocks with intelligent insights, comprehensive financial data, and AI-driven research—all in one platform.

Comment

Inspiration

Stock market research often involves switching between multiple platforms to understand a company's financial performance, valuation, historical trends, and latest developments. Platforms like Screener.in make fundamental analysis easier, but I wanted to take the experience further by bringing Indian and US stocks together in a single AI-powered research platform.

That's how Screen-Finance started.

The idea was to combine the simplicity of traditional stock screeners with the intelligence of Agentic AI, allowing investors to move beyond simply looking at financial numbers and actually understand the story behind them.

My goal is to make financial research more accessible, insightful, and efficient for everyone, from beginners to experienced investors.

What It Does

Screen-Finance is an AI-powered stock research and financial analysis platform designed for Indian and US equity markets.

The platform is designed to provide a comprehensive view of publicly listed companies through an intuitive, data-rich dashboard.

Its core features include:

  • Unified Stock Research: Explore companies listed across Indian and US stock exchanges.
  • Financial Dashboards: Analyze revenue, profitability, valuation ratios, quarterly results, balance sheets, and cash flows.
  • Interactive Stock Charts: Understand historical price movements and market trends.
  • Company Comparisons: Evaluate multiple companies using consistent financial metrics.
  • Intelligent Stock Screening: Discover companies based on financial filters and investment criteria.
  • AI-Powered Research: Use intelligent agents to interpret financial information, investigate company performance, and generate understandable insights.
  • Document Intelligence: Analyze annual reports, earnings transcripts, and regulatory filings through Retrieval-Augmented Generation (RAG).

The long-term vision is to make Screen-Finance more than a financial dashboard. I want it to become an intelligent research assistant that helps users make better-informed investment decisions.

How We Built It

I approached Screen-Finance with a modular, scalable architecture, separating the frontend experience, financial data processing, and AI research capabilities.

Frontend — React Ecosystem

The frontend is built around React, with a responsive, dark-themed financial dashboard inspired by professional stock research platforms. The interface focuses on structured financial tables, interactive charts, company metrics, and intuitive navigation.

Backend — Python and FastAPI

FastAPI is the chosen backend framework for handling stock information, financial calculations, API integrations, authentication, and communication between the frontend and AI services.

Database — PostgreSQL and Redis

PostgreSQL is intended to manage structured financial information, company profiles, historical records, user accounts, and watchlists. Redis provides a caching layer to reduce repeated API requests and improve response times.

Agentic AI — LangGraph and MCP

The AI architecture is designed around LangGraph to coordinate specialized research agents responsible for financial analysis, company comparisons, document research, and risk identification.

Model Context Protocol (MCP) provides a standardized approach for connecting AI agents with financial data services, databases, and retrieval tools.

RAG — Qdrant Vector Database

Qdrant serves as the planned vector database for retrieving relevant information from annual reports, earnings transcripts, and financial documents. This allows AI-generated explanations to be grounded in retrieved evidence rather than relying entirely on an LLM's existing knowledge.

Infrastructure — Docker

Docker and Docker Compose are part of the deployment strategy to simplify development, manage dependencies, and support consistent environments.

The project is being developed incrementally, starting with the user interface and expanding toward live financial data integration and intelligent multi-agent workflows.

Challenges We Ran Into

Building Screen-Finance has introduced several interesting engineering challenges.

1. Designing a Data-Heavy Financial Interface

One of the first challenges was creating a dashboard capable of displaying large financial tables, historical information, and interactive charts without making the interface overwhelming. I spent considerable effort improving the layout, navigation, visual consistency, and overall user experience.

2. Supporting Two Different Stock Markets

Indian and US financial markets have different reporting standards, currencies, financial periods, and data providers. Designing a unified platform requires careful normalization so users can analyze companies consistently without losing important market-specific details.

3. Financial Data Availability and Accuracy

Finding reliable financial data is more complicated than simply calling an API. Different providers offer different levels of coverage, freshness, and historical information. Building a dependable data pipeline while managing API limitations remains an important part of the development process.

4. Designing Reliable AI Research Workflows

Financial analysis requires accuracy and transparency. An AI agent must distinguish between verified financial facts, calculated metrics, and its own interpretation. This makes source attribution, structured tool access, retrieval quality, and hallucination prevention essential design considerations.

5. Balancing Scalability and Development Cost

Since the project is being developed at an early stage, another challenge is choosing infrastructure that is affordable today but can support additional users, larger datasets, and more complex AI workflows later.

These challenges have helped shape the technical decisions and development roadmap.

What's Next

Screen-Finance is an evolving project, and the next phase is focused on turning the improved frontend into a fully functional AI-powered financial research platform.

The upcoming milestones include:

  1. Live Market Data Integration: Connect reliable financial data providers for Indian and US stocks.
  2. Complete Backend Integration: Build FastAPI endpoints for company information, financial statements, ratios, and historical stock prices.
  3. User Authentication and Watchlists: Introduce secure login, personalized dashboards, and saved stock lists.
  4. RAG-Powered Financial Research: Enable users to ask questions about annual reports, earnings calls, and company filings.
  5. Multi-Agent AI Analysis: Implement specialized LangGraph agents that collaborate to produce structured company research reports.
  6. Natural Language Stock Screening: Allow users to search using prompts such as "Find profitable Indian technology companies with low debt and consistent revenue growth."
  7. Portfolio Intelligence: Introduce portfolio analysis, risk insights, and personalized research alerts.
  8. Cloud Deployment: Build a reliable deployment pipeline with Docker, automated testing, monitoring, and scalable infrastructure.

Final Vision

Screen-Finance aims to bridge the gap between traditional stock screening and intelligent financial research.

Instead of spending hours gathering financial information across different websites, users should be able to explore companies, understand financial performance, compare opportunities, and investigate investment risks from one platform.

One platform. Two markets. Smarter research with Agentic AI.

Built With

Share this project:

Updates

Submission history