← Engineering
Completed2024–2025

Conversational Agents Course

Seven practical workshops on building, exposing and evaluating conversational agents.

Nathan Rihet teaching a university class about generative AI integration with LangChain
Teaching applied generative AI and LangChain
Seven-workshop syllabusFrom language-model foundations to a working conversational-agent demonstration
  1. 01LLMs
  2. 02REST
  3. 03FastAPI
  4. 04Chains
  5. 05Tools
  6. 06Evaluate
  7. 07Demo

Problem

Project brief

Give Master’s students enough structure and hands-on practice to build and assess a complete API-based conversational agent.

Contribution

Role and scope

Designed the curriculum, taught the workshops and built the reference implementations.

Python · FastAPI · LangChain · LLMs

Implementation

Key technical decisions

  1. Structured the material from REST and FastAPI fundamentals through tools, function calling and evaluation.
  2. Separated routes, Pydantic models and the LLM service in the reference application.
  3. Finished with group projects and demonstrations so students had to combine the complete workflow.

Constraints

Engineering constraints

  • Fit API fundamentals, LLM orchestration and evaluation into seven progressive workshops.
  • Keep the reference implementation approachable while preserving a realistic service structure.
  • Provide exercises that connect theory to observable application behavior.

Outcome

Result

More than 20 Master’s students completed the seven-workshop sequence and presented working group projects.

View source