AI Stock Insights: Technical Analysis, LLM Generation & Telegram Publishing

This workflow runs twice daily on a schedule, processing a predefined list of stock tickers like GAZP, SBER, and LKOH. It fetches historical market data via a broker API, computes key technical indicators including RSI, EMA/SMA, MACD, Bollinger Bands, and ADX using custom functions. An AI model (OpenAI or OpenRouter) then generates structured investment posts with titles and summaries based on the analysis.

Results are stored in a PostgreSQL database for historical tracking and analytics. A draft post is sent to Telegram with inline buttons for 'Publish' or 'Retry', enabling human-in-the-loop moderation. Telegram triggers handle button callbacks: publishing to the final channel or re-running the generation process. Robust error handling covers missing data or invalid AI outputs.

Ideal for fintech startups, investment bloggers, and analysts automating content creation. Benefits include multi-ticker efficiency, flexible extensibility for new indicators/channels, and time-saving automation of market analysis and social publishing, reducing manual effort from hours to minutes daily.