Indonesian stock market research agent with Sectors App, Gemini & Tavily

This n8n workflow demonstrate how to configure AI Agent for financial research purposes especially for IDX data through Sectors App API.

use cases:

• research stock market in Indonesia.
• analyze the performance of companies belonging to certain subsectors or company
• comparing financial metrics between BBCA and BBRI
• providing technical analysis for certain ticker stock movement
• and many more

all from conversational agent UI chat.

Main components

• Input-n8nChatNative: handling and process input from native n8n chat ui
• Input-TelegramBot: handling and process input from Telegram Bot
• Input-WebUI(Webhook): handling and process input from hosted Web UI through webhook
• Main Agent: processing raw user queries and delegate task to specialized agent if needed.
• Spec Agent - Sectors App: make request to Sectors App API to get real time financial data listed in IDX from available endpoint
• Spec Agent - Web Search: make web search from Google Grounding (Gemini API) and Tavily Search.
• Vector Document Processing: process document upload from user into embedding and vector store.

How it works

• user queries may received from multiple platform (we use three here: Telegram, hosted Web UI, and native n8n chat UI)
• if user also uploading document, it will process the document and store it in vector store
• the request send to the Main Agent to process the queries
• the Main Agent decide the task to delegate to Specialized Agent if nedded.
• the result then sent back to user based on the platform

How to use

You need this API:

• Gemini API: get it free from https://aistudio.google.com/
• Tavily API: get it free from https://www.tavily.com/
• Sectors App API: get it from https://sectors.app/api/

you can optionally change the model or adding fallback model to handle token request, cause it may use quite many tokens.