Implementing a Generative AI-Powered Assistant for Student Support​

Our client operates an edtech platform featuring a marketplace where schools can explore and purchase a diverse range of courses offered on the platform​

Challenge​

For On- Demand courses Students often experience long wait times for their questions to be answered, which can impede their learning and productivity. To address this issue, they wanted us to implement a smart assistant of automated system using generative artificial intelligence (AI) to provide timely and accurate responses to student inquiries. The goal is to enhance the efficiency of the answering process, reduce wait times, and improve overall student satisfaction​

Solution

Implemented a Gen AI-driven automated system designed to provide prompt and accurate responses to student inquiries. This system was developed to address common questions and issues while ensuring that more complex queries could be seamlessly escalated to human staff if necessary.

• Designed a LLM powered assistant with the dynamic prompting based on the entities extracted from the student queries using the SpaCy/Hugging Face models.​
• Implemented a RAG based solution utilizing advanced RAG technique like Context Enrichment Techniques by embedding individual sentences and extending context.​
• System is integrated with a comprehensive knowledge base, which includes information about academic programs, administrative procedures, and frequently asked questions (FAQs).​
• Students can provide feedback on the quality of responses received. This feedback is used to continuously refine and improve the AI system’s accuracy and effectiveness.​
• The system includes analytics tools that track query types, response times, and user satisfaction. This data helps in monitoring performance and identifying areas for further improvement.​
• The system includes an escalation protocol that directs the query to the appropriate human staff member for resolution that AI cannot address​