Skip to content
Data EngineeringIn ProgressJan 2026

Automated Gold Price Tracking System

Real-time gold price monitoring with automated scraping, API integration, data storage in Google Sheets, Looker Studio visualization, and Telegram bot alerts.

Role
Data Engineer (Personal Project)
Timeline
2026 – Ongoing
Team
Solo
Status
In Progress

The problem

Checking gold prices by hand across several sources is tedious and easy to forget. I wanted prices captured on a schedule, stored, charted, and flagged only when something meaningful happens.

Constraints

  • Zero infrastructure to run or pay for
  • Alerts must stay signal, not noise
  • Almost no maintenance once it is running

Architecture

  1. Step 1, Ingest

    Apps Script triggers

    Scheduled and serverless

  2. Step 2, External

    Price APIs

    Gold price sources

  3. Step 3, Data

    Google Sheets

    Normalised, de-duplicated store

  4. Step 4, Interface

    Looker Studio

    Always-fresh trend dashboard

  5. Step 5, Interface

    Telegram bot

    Threshold alerts

Time-driven Google Apps Script triggers pull gold prices from APIs, normalise and de-duplicate them into Google Sheets, feed an always-fresh Looker Studio dashboard, and send Telegram alerts when prices cross configurable thresholds.

Engineering decisions

  1. Serverless triggers instead of a server

    Context
    A personal data project dies the moment it needs babysitting.
    Decision
    Ran the entire pipeline on Google Apps Script time-driven triggers.
    Trade-off
    Tied to the Apps Script runtime and its quotas.
  2. A spreadsheet as the database

    Context
    The data had to be easy to inspect and chart without extra services.
    Decision
    Stored normalised, de-duplicated rows in Google Sheets, which Looker Studio reads directly.
    Trade-off
    Not a data warehouse — but for one time series it keeps everything visible and free.
  3. Threshold-based alerts

    Context
    Alerts on every price change would quickly be ignored.
    Decision
    The Telegram bot fires only when prices cross configurable thresholds.

Outcomes

Fully Automated

Automation

Telegram Bot

Alert System

Looker Studio

Visualization

  • Runs fully automated on a schedule
  • Always-fresh dashboards straight from the data store
  • Costs nothing to run and needs almost no maintenance

Reflections

Lessons learned

More than anything, it proved how much of a superpower serverless, schedule-driven automation is for personal data projects — the entire system costs nothing to run and needs almost no maintenance once the triggers are set, which is exactly what keeps it alive long after the initial build.

What's next

It runs fully automated today and remains an ongoing personal project I extend whenever a new data source becomes useful.

Proof — verify it yourself

The source and demo for this project aren’t public. Get in touch for a walkthrough.

Tech stack

  • Google Apps Script
  • JavaScript
  • APIs
  • Looker Studio
  • Telegram Bot

Work with me

Facing a similar challenge?

I build systems like this end to end — from the first architecture sketch to production. Tell me about yours.