Meraki Intelligence

01 / The Business Challenge

A wholesale gemstone business sourcing internationally may be affected by changes in trade rules, exchange rates, shipping conditions, and gemstone availability. Keeping up with these developments manually requires time and attention.

I started this project independently with Meraki Gemstones in mind. I have not yet pitched it to the business.

02 / My Solution

I developed Meraki Intelligence: a functioning, automated business intelligence email system that collects news and public data, filters headlines for potential business relevance, tracks exchange rates, and prepares a daily HTML email.

I used JavaScript rules and keyword matching to filter and prioritize headlines. The current system does not use a generative AI model or machine learning.

03 / How the System Works

A Google Apps Script time-driven trigger starts the daily workflow. The report is designed to run around 8:00 AM Central Time; trigger execution and email delivery are approximate, not guaranteed at precisely 8:00 AM.

The script reads configuration from Google Sheets, retrieves Google News RSS and other public RSS feeds, applies relevance rules, and suppresses repeated or previously emailed stories. It retrieves exchange-rate data from the Frankfurter API and compares it with previously recorded observations.

The selected stories, source links, dates, review suggestions, and currency information are assembled into an HTML email delivered through Google MailApp. Google Sheets stores monitoring data and report history.

India, Brazil, China, Madagascar, and Russia are provisional monitoring countries. These are monitoring assumptions, not confirmed countries from which Meraki currently purchases inventory. The system does not track Meraki’s actual suppliers.

04 / Features and Functionality

Automated news monitoring

Retrieves headlines and public updates from RSS feeds.

Business relevance filtering

Uses rules and keyword matching to prioritize gemstone industry developments, trade changes, supply issues, and major logistics disruptions.

Currency monitoring

Retrieves exchange-rate data for monitored currencies and compares rates against previously recorded observations.

Daily email briefing

Formats selected headlines, sources, dates, relevant review suggestions, and currency information into an accessible HTML email.

Duplicate detection

Identifies and suppresses repeated or previously emailed stories.

Source monitoring

Records feed availability and flags incomplete data coverage.

Scheduling

Runs through an automated daily time-driven trigger, with approximate morning delivery.

Testing and logging

Includes report previews, test emails, execution history, and email authorization safeguards.

05 / Technology Stack

Google Apps Script and JavaScript

Automation, rules-based filtering, and report assembly.

Google Sheets

Configuration, monitoring observations, and report history.

Google News RSS and public RSS feeds

News and public-update collection.

Frankfurter API

International exchange-rate data.

Google MailApp

HTML email delivery.

Apps Script time-driven triggers

Daily scheduling.

06 / Business Value

I designed the system to reduce the need to search multiple news sources manually, improve awareness of international developments, make currency information accessible, and highlight potential sourcing and operational risks.

Using familiar Google tools and rules-based processing creates a practical, cost-conscious daily information workflow without a generative-model dependency. These are intended benefits; cost reductions, financial returns, purchasing improvements, and productivity savings have not been documented.

The briefing collects and organizes information for review. It does not predict disruptions, calculate profit impact, guarantee real-time alerts, independently verify all news, or determine whether an import is legally permitted.

07 / What I Learned

Building this system reinforced the importance of matching automation to a specific business need. A focused daily briefing can be useful without adding a generative AI model.

I learned to treat relevance rules, duplicate handling, source availability, and historical observations as core parts of the reporting workflow. Clear source links, visible coverage gaps, testing, and authorization safeguards help make an automated report easier to review responsibly.