Project: [Reasoning Agents] - 4C Hair Products Agent

[kamsiileagu26]

Track

Reasoning Agents (Azure AI Foundry)

Project Name

4C Hair Products Agent

GitHub Username

kamsiileagu26

Repository URL

https://github.com/kamsiileagu26/4c-hair-products-agent

Project Description

An ETL agent that automatically scrapes the internet for 4C hair product data, extracts structured information using Microsoft Phi-4-mini-instruct on Azure AI Foundry (Foundry IQ), and stores it in a Supabase database. The agent collected 80+ real product data points across 10 categories including shampoo, conditioner, oil, gel, butter and more. Built with TypeScript and GitHub Copilot.

Demo Video or Screenshots

https://drive.google.com/file/d/1UYsvGQKlWSkIyzU3PVR7WbBsICkgOh6Q/view?usp=sharing

Primary Programming Language

TypeScript/JavaScript

Key Technologies Used

• Azure AI Foundry (Foundry IQ)
• Phi-4-mini-instruct (Microsoft)
• Firecrawl
• Supabase
• TypeScript
• Claude.ai
• GitHub Copilot

Submission Type

Individual

Team Members

No response

Submission Requirements

• My project meets the track-specific challenge requirements
• My repository includes a comprehensive README.md with setup instructions
• My code does not contain hardcoded API keys or secrets
• I have included demo materials (video or screenshots)
• My project is my own work with proper attribution for any third-party code
• I agree to the Code of Conduct
• I have read and agree to the Disclaimer
• My submission does NOT contain any confidential, proprietary, or sensitive information
• I confirm I have the rights to submit this content and grant the necessary licenses

Quick Setup Summary

1. Clone the repo: git clone https://github.com/kamsiileagu26/4c-hair-products-agent
2. Install dependencies: npm install
3. Add .env file with: FIRECRAWL_API_KEY, AZURE_FOUNDRY_ENDPOINT, AZURE_FOUNDRY_KEY, SUPABASE_URL, SUPABASE_KEY
4. Run: npx ts-node src/agent.ts

Technical Highlights

• Built a full Extract, Transform, Load pipeline in TypeScript that runs automatically end to end
• Integrated Microsoft Phi-4-mini-instruct via Azure AI Foundry to extract structured data from unstructured web content
• Implemented smart filtering to skip invalid rows — only saves rows with both hair_type and product_name
• Used timeout handling to skip slow requests and keep the pipeline running
• Collected 80+ real 4C hair product data points across 10 categories

Challenges & Learnings

• Getting Azure AI Foundry quota approved for new accounts was a challenge — solved by using Microsoft's Phi-4-mini-instruct model which had available quota
• Phi-4-reasoning model ignored JSON formatting instructions so switched to Phi-4-mini-instruct which followed instructions reliably
• Many sites like Reddit, TikTok and Pinterest block scraping — solved with try/catch error handling to skip blocked URLs and keep the pipeline running
• Learned how to build a full ETL pipeline from scratch using TypeScript and Azure AI Foundry

Contact Information

kamsiileagu36@gmail.com

Country/Region

united Kingdom