The Bot-om Line: Agentic AI for Customer Service

Objective and Focus

The workshop will be a walkthrough on how to create and implement your own (Agentic RAG) LLM for customer service. We will create a system around the LLM that takes a prompt from the customer and decides whether it can answer the response or if it has to forward the customer to staff. It will also incorporate a vector database of information from the “company” that will facilitate it in answering any questions.

Structure and Content

• Open Google Colab
• Import requirements (pip install torch, transformers)
• Torch
• Transformers
• Faiss
• Download the open source LLM (seq2seq)
• BERT
• Set up an agentic workflow
• Create the router (Decides if the prompt can be answered or not)
• Create the responder (Writes the actual response with RAG)
• Implement RAG (Retrieval Augmented Generation) to retrieve relevant information regarding the prompt and feed it into the model
• Vector database
• DONE!!!

Style and Approach

Interactive and practical, with a focus on hands-on learning.

Target Audience

This workshop is designed for:

• AI Enthusiasts: Individuals curious about the inner workings of AI agents and their real-world applications.
• Developers and Engineers: Those interested in building AI-driven systems or exploring the mechanics behind conversational and task-oriented AI agents.
• Researchers and Students: Aspiring or experienced researchers eager to deepen their understanding of AI agent architectures and implementations.
• Product Designers and Managers: Professionals seeking insights into how AI agents can enhance user experiences and streamline processes.

Required Preparations

• A laptop
• Wifi connection
• Web browser
• Basic Python knowledge and/or tech savviness

Host Information

Shariq Sayied Ali is the Chief Technological Officer and co-founder of M10 AI. At 19, Shariq is a distinguished alumnus and teaching assistant at Chalmers University of Technology. With seven years of programming experience Shariq is already an experienced developer and a rising star in the AI field. He has led numerous workshops and guided participants in implementing state-of-the-art AI models. His dedication to studying and applying the latest research ensures he remains at the forefront of advancements in artificial intelligence.