Egypt Gold Bot: AI Chatbot Integration Service in Action

For residents in Egypt, tracking the local price of gold in EGP is a high-stakes, high-stress activity. The market is volatile, and reliable data is scattered across various Arabic-language websites.

Making an informed decision about when to buy, sell, or hold requires constant manual checking and an attempt to "feel" the market trend. This manual process is time-consuming, prone to error, and lacks a data-backed foundation, leading to missed opportunities or costly mistakes.

The client needed a way to automate this entire process and receive proactive, intelligent advice. This is a classic scenario where a custom chatbot integration can provide immense value by turning data chaos into clarity.

I developed Egypt Gold Bot, a comprehensive, two-part solution that goes far beyond a simple chatbot. It’s a complete data-to-decision pipeline that I built from the ground up.

Part 1: Building a Robust Gold Price Data Engine

Before an AI can give advice, it needs reliable data. I built a custom data pipeline from the ground up:

• Automated Scraper: I developed a Python-based scraper to fetch prices from the target Arabic website. I hardened it with proxy rotation, random user-agents, and robust retry logic to ensure 24/7 operation.

• Intelligent Data Consolidation: The system doesn't just store raw data. I designed it to automatically process and consolidate data at different intervals:

  • Real-time: It captures live prices for all carats, weights (troy ounce, gold pound), and the international USD price every 15 minutes.

  • Hourly: It automatically calculates the min, max, and average prices for each hour, providing a view of intra-day volatility.

  • Daily: It rolls up the hourly data to create a powerful daily summary, calculating opening_price_sell, closing_price_sell, and a trend_indicator (increased, decreased, stable).

• Dedicated API: I built a FastAPI to expose this rich, consolidated data via secure endpoints, allowing me to retrieve historical trends with simple filters.

Part 2: An AI Chatbot Integration Service for Smart Advice

With a reliable data source, I implemented the ai chatbot integration service using n8n to connect my new API, OpenAI, and Telegram. This wasn't a simple chatgpt assistant integration; it was a dual-mode system I built for proactive insights.

1. Mode 1: The Proactive Daily Advisor As seen in the n8n workflow, a daily schedule trigger kicks off a process. It calls my API to fetch the current 21-carat price and the full 365-day daily history. This rich data is fed to a specialized "Gold Advisor" AI prompt, which analyzes the long-term trend, the recent momentum, and the current price to generate a clear "Buy, Sell, or Hold" recommendation and sends it to the user's Telegram.

   

2. Mode 2: The On-Demand Q&A Assistant This is the ai virtual assistant setup. When the user sends a message, the Telegram trigger activates. The exact same data-fetching process occurs, feeding the 365-day history to a general-purpose "Assistant" prompt. This allows the user to ask specific questions (e.g., "What was the trend last month?") and receive an data-backed, conversational answer. This provides a 24/7 ai customer support chatbot experience for their personal finances.

   

The EGP Gold Bot project transformed a stressful, manual task into a fully automated, intelligent advisory service. The user no longer needs to check websites; they receive a proactive daily briefing with a clear recommendation. This solution is a powerful example of how my integration skills can turn a stream of raw data into actionable, automated intelligence.

This project combined robust data scraping and API development with a flexible Custom AI Assistant Integration, creating a service that empowers the user to make confident, data-driven financial decisions.