User Satisfaction Analyst Agent
by Carlos Pumar-Frohberg
Carlos Pumar-Frohberg's agent analyzes client satisfaction using Stack Exchange data, focusing on UI discussions to find frustration patterns. An orchestrator agent routes questions to a MongoDB agent for 'what'/'how' queries or a Cipher agent that translates natural language into Neo4j graph queries, often calling both for safety.
Project from the first cohort of the AI Engineering Buildcamp, by Carlos Pumar-Frohberg.
The architecture uses a Docker pipeline to fetch Stack Exchange data and dump it into two stores: MongoDB for unstructured data and Neo4j for graph data. An orchestrator agent decides where to route user questions: if the question is about "what" or "how," it goes to the MongoDB agent; if it's about relationships, it goes to a "Cipher" agent that translates natural language into graph queries. Carlos noted that the orchestrator often calls both agents simultaneously to be on the safe side.
Tech stack: Docker, MongoDB, Neo4j, orchestrator agent, Cipher (NL → Cypher).