Design and implement alangchain based system that classifies incoming emails into three categories and executes specific actions based on the category. The system should enhance customer interaction by providing automated responses tailored to inquiries, reviews, and assistance requests.
- Database: Develop an SQL database to store details about film equipment.
- Action: When an inquiry email is received, check the database for item
availability.
- If available, reply with the item's price.
- If not available, suggest similar available items.
- Positive Reviews: Thank the sender and encourage them to share their experience on social media.
- Negative Reviews: Escalate to the CRM system for follow-up with a phone call from customer service and offer a gift voucher in the reply.
- Documentation: Create a document with frequently asked questions about film equipment using ChatGPT or from equipment manuals.
- RAG Pipeline: Use a Retrieval-Augmented Generation approach to search for
solutions to reported equipment issues. (As of now, Not working completely.)
- If a suitable solution is found, provide it in the reply.
- If no solution is found, escalate the issue to customer service.
- Forwarding: Emails that do not fit into the above categories should be forwarded to customer service for further evaluation.
As of now only one relation is considered as we are dealing only with the category, quantity, rental price details of an equipment.