Full-Stack Development for AI-Powered News Sentiment & Trading Signal Platform
Budget: $250 – $750 USD
Project Overview:
We are seeking developer to build an end-to-end solution that processes financial news in real time, analyzes sentiment using GPT, and generates actionable trading signals. The system should combine web scraping, natural language processing, trading signal generation, and financial data integration into a single robust platform with a front-end dashboard and API access.
Key Deliverables:
1. Web Scraping & Data Ingestion
○ Develop a scalable web crawler capable of scraping and aggregating news from multiple financial and general news websites.
○ Handle anti-bot measures, ensure reliability, and provide options for periodic or real-time scraping. Support logging, java-script, clicking on buttons like "continue reading"
○ Maintain historical records of scraped articles for backtesting.
○ Create also an option to connect to APIs like FinnHub
2. News Processing & Analysis (with GPT)
○ Use GPT or LLMs to evaluate sentiment, relevance, and potential impact of news articles.
○ Classify and rate news on a confidence scale (e.g., positive, negative, neutral impact on a given ticker).
○ Generate trading recommendations (e.g., ticker → buy/sell/hold + confidence ratio).
○ To shall be well-optimized to minimize the token usage
3. Market Data Integration
○ Connect with Interactive Brokers API for:
§ Live price feeds
§ Historical market data
§ Storing and updating records for as many tickers as possible.
4. Signal Generation & Storage
○ Aggregate GPT-based analysis with real-time market data.
○ Generate structured trading signals with metadata (timestamp, ticker, sentiment, action, confidence).
○ Store signals in a database for backtesting, performance evaluation, and historical insights.
5. Front-End Dashboard
○ Develop a modern, user-friendly web interface to present:
§ General market sentiment summary.
§ Company-specific news sentiment breakdown.
§ Recently generated trading signals with filtering/searching.
6. API Access
○ Expose a REST (or GraphQL) API for external systems to subscribe and receive trading signals in real-time.
○ Include authentication and access controls for external subscribers.
Deliverables & Milestones:
1. Initial web crawler setup with news ingestion pipeline.
2. GPT sentiment analysis module with rating and ticker mapping.
3. Interactive Brokers integration with live market data.
4. Signal generation and storage system.
5. Front-end dashboard.
6. API exposure for trading signals.
Tech stack (proposed)
• Backend / API:
• Python (FastAPI or Flask for APIs, Celery/RQ for tasks)
• LangChain / OpenAI API for GPT integration
• Interactive Brokers API (IBKR TWS / IB Gateway)
• Web Scraping & Processing:
• Python (Scrapy, Playwright, or BeautifulSoup for scraping)
• Async frameworks (aiohttp, asyncio) for scalability
• Database & Storage:
• PostgreSQL (structured data)
• MongoDB (unstructured article storage)
• Qdrant (if needed for GPT)
• Front-End:
• React
• Infrastructure / Deployment:
* Docker
We are seeking developer to build an end-to-end solution that processes financial news in real time, analyzes sentiment using GPT, and generates actionable trading signals. The system should combine web scraping, natural language processing, trading signal generation, and financial data integration into a single robust platform with a front-end dashboard and API access.
Key Deliverables:
1. Web Scraping & Data Ingestion
○ Develop a scalable web crawler capable of scraping and aggregating news from multiple financial and general news websites.
○ Handle anti-bot measures, ensure reliability, and provide options for periodic or real-time scraping. Support logging, java-script, clicking on buttons like "continue reading"
○ Maintain historical records of scraped articles for backtesting.
○ Create also an option to connect to APIs like FinnHub
2. News Processing & Analysis (with GPT)
○ Use GPT or LLMs to evaluate sentiment, relevance, and potential impact of news articles.
○ Classify and rate news on a confidence scale (e.g., positive, negative, neutral impact on a given ticker).
○ Generate trading recommendations (e.g., ticker → buy/sell/hold + confidence ratio).
○ To shall be well-optimized to minimize the token usage
3. Market Data Integration
○ Connect with Interactive Brokers API for:
§ Live price feeds
§ Historical market data
§ Storing and updating records for as many tickers as possible.
4. Signal Generation & Storage
○ Aggregate GPT-based analysis with real-time market data.
○ Generate structured trading signals with metadata (timestamp, ticker, sentiment, action, confidence).
○ Store signals in a database for backtesting, performance evaluation, and historical insights.
5. Front-End Dashboard
○ Develop a modern, user-friendly web interface to present:
§ General market sentiment summary.
§ Company-specific news sentiment breakdown.
§ Recently generated trading signals with filtering/searching.
6. API Access
○ Expose a REST (or GraphQL) API for external systems to subscribe and receive trading signals in real-time.
○ Include authentication and access controls for external subscribers.
Deliverables & Milestones:
1. Initial web crawler setup with news ingestion pipeline.
2. GPT sentiment analysis module with rating and ticker mapping.
3. Interactive Brokers integration with live market data.
4. Signal generation and storage system.
5. Front-end dashboard.
6. API exposure for trading signals.
Tech stack (proposed)
• Backend / API:
• Python (FastAPI or Flask for APIs, Celery/RQ for tasks)
• LangChain / OpenAI API for GPT integration
• Interactive Brokers API (IBKR TWS / IB Gateway)
• Web Scraping & Processing:
• Python (Scrapy, Playwright, or BeautifulSoup for scraping)
• Async frameworks (aiohttp, asyncio) for scalability
• Database & Storage:
• PostgreSQL (structured data)
• MongoDB (unstructured article storage)
• Qdrant (if needed for GPT)
• Front-End:
• React
• Infrastructure / Deployment:
* Docker