Azure AI Bilingual Chatbot with Web Scraping
Budget: $750 – $1,500 USD
Title: Microsoft Azure Generative AI Chatbot with RAG, Web Scraping & Semantic Search (English & Arabic)
Description:
Design and implement a Retrieval-Augmented Generation (RAG) solution to power an intelligent contextual based chatbot for our website.
Project Overview:
1. Web Scraping – Bilingual Support
Develop and maintain a robust scraping pipeline to extract data from both the English and Arabic versions of the Our website, ensuring language-specific nuances are accurately captured and enhanced error detection and handling logic across all stages of the pipeline - including web scraping, content parsing, and query response generation - to improve system resilience, minimize downtime, and ensure reliable end-to-end performance.
2. Optimized OCR Workflow for PDFs and Images
Implement a refined Optical Character Recognition (OCR) pipeline tailored for processing both documents (like PDF, Word & Excel) and image-based content, integrated with HTML parsing mechanisms to ensure consistency and alignment across diverse document formats.
3. Real-Time Website Monitoring
Implement continuous monitoring mechanisms to detect and capture website updates, dynamic content changes, and the addition of new pages in real time.
4. Data Structuring for Vector Indexing
Transform the scraped content into a well-structured, machine-readable format optimized for ingestion into Axure’s vector search infrastructure, ensuring semantic consistency across both languages.
5. WebSocket API Development
Design and deploy a scalable WebSocket-based API to serve the RAG solution, enabling real-time interaction between the front-end chatbot and the backend knowledge retrieval system.
6. Multilingual Query Handling & Response Generation
Enable the WebSocket to accept and process user queries in both English and Arabic, leveraging the RAG architecture to deliver accurate, contextually relevant responses based on the most up-to-date website content.
7. Testing
Test and validate the chatbot across multiple scenarios, ensuring a bug-free, production-ready deployment.
Requirements:
• Proven experience building RAG-based architectures with LLM integration in Azure
• Experience with web scraping tools
• Familiarity with vector databases in Azure
• Strong grasp of semantic search, embedding models, and multilingual NLP.
• Proficiency in Python and LLM frameworks used in Azure
• Ability to support both English and Arabic NLP tasks effectively.
• Good understanding of scalable architecture and performance optimization.
Description:
Design and implement a Retrieval-Augmented Generation (RAG) solution to power an intelligent contextual based chatbot for our website.
Project Overview:
1. Web Scraping – Bilingual Support
Develop and maintain a robust scraping pipeline to extract data from both the English and Arabic versions of the Our website, ensuring language-specific nuances are accurately captured and enhanced error detection and handling logic across all stages of the pipeline - including web scraping, content parsing, and query response generation - to improve system resilience, minimize downtime, and ensure reliable end-to-end performance.
2. Optimized OCR Workflow for PDFs and Images
Implement a refined Optical Character Recognition (OCR) pipeline tailored for processing both documents (like PDF, Word & Excel) and image-based content, integrated with HTML parsing mechanisms to ensure consistency and alignment across diverse document formats.
3. Real-Time Website Monitoring
Implement continuous monitoring mechanisms to detect and capture website updates, dynamic content changes, and the addition of new pages in real time.
4. Data Structuring for Vector Indexing
Transform the scraped content into a well-structured, machine-readable format optimized for ingestion into Axure’s vector search infrastructure, ensuring semantic consistency across both languages.
5. WebSocket API Development
Design and deploy a scalable WebSocket-based API to serve the RAG solution, enabling real-time interaction between the front-end chatbot and the backend knowledge retrieval system.
6. Multilingual Query Handling & Response Generation
Enable the WebSocket to accept and process user queries in both English and Arabic, leveraging the RAG architecture to deliver accurate, contextually relevant responses based on the most up-to-date website content.
7. Testing
Test and validate the chatbot across multiple scenarios, ensuring a bug-free, production-ready deployment.
Requirements:
• Proven experience building RAG-based architectures with LLM integration in Azure
• Experience with web scraping tools
• Familiarity with vector databases in Azure
• Strong grasp of semantic search, embedding models, and multilingual NLP.
• Proficiency in Python and LLM frameworks used in Azure
• Ability to support both English and Arabic NLP tasks effectively.
• Good understanding of scalable architecture and performance optimization.