Algorithmic trading engine

React Type Script SQL Python Postgre
FinTech FinTech

An algorithmic trading engine designed to help users monitor market activity, analyze asset performance, and review trade signals through a single interface. It supports multiple asset categories, including FX, stocks, commodities, and crypto, and combines chart-based analysis, structured trade history, and real-time signal delivery with market context gathered from multiple financial news sources and interpreted through AI-powered trade insights. 

hero image

Project request

The project centered on implementing the data, insight, and delivery layers required for a trading analytics product. This included collecting information from external financial news sources, generating AI-based explanations for signals and market changes, and making those insights available through the platform, subscription flows, and notification channels. 

Key features: 

  • aggregation of information from multiple financial news sources 
  • AI-powered interpretation of trade signals and market behavior 
  • multi-market support across FX, stocks, commodities, and crypto 
  • asset and currency pair navigation 
  • historical charting with timeframe filters 
  • area and candle chart modes 
  • a Trades section with action, dates, result, percentage change, and reason 
  • real-time in-app signal streaming 
  • delayed access for free users 
  • premium subscription access 
  • browser, email, and Telegram alerts 

Screenshots

Our approach

We built the platform around an asynchronous backend that collects market context from multiple financial news sources, processes it through a hybrid AI pipeline, and delivers validated signals to the product layer. 

 

The analytical flow combines LLM-based interpretation, RAG-based retrieval of historical patterns, and XGBoost-based prediction to generate trade explanations with stronger context. To improve reliability, the backend is split into isolated services for news collection, news analysis, signal validation, trade execution, and API delivery. 

 

Signals are filtered through rule-based checks, enriched with historical context, and passed to an execution layer integrated with Interactive Brokers, where the system handles position sizing, bracket orders, reverse-signal logic, and trade state reconciliation. 

On the user-facing side, these insights are offered within the web app as real-time signals, structured trade data, premium content, and alerts delivered via browser, email, and Telegram. 

 

  • news aggregation from multiple financial news sources 
  • hybrid AI analysis with LLM, RAG, and XGBoost 
  • isolated services for collection, analysis, validation, execution, and API delivery 
  • Interactive Brokers integration for execution and position management 
  • risk control logic, including bracket orders, reverse-signal handling, and reconciliation 
  • real-time delivery with browser, email, and Telegram notifications 

Client’s feedback

Viktor Krig Incorporations Team Lead at Online Taxman

bART Solutions played a key role in enhancing our tax advisory services with their AI-driven platform. Their expertise in AI and secure cloud infrastructure, along with their user-focused design, has greatly improved our efficiency and client satisfaction. We have seen a significant reduction in processing times and errors. Their professional and thorough approach to the entire development process has exceeded our expectations. We look forward to continuing our partnership and achieving even greater milestones together.

5.0
Quality 5.0
Cost 5.0
Schedule 5.0
Willing to Refer 5.0