Development of a foundational AI chatbot named Sevi for a fictional airline, CloudSeven Airlines, designed as a customer service assistant. The chatbot uses a local LLM (Qwen 2.5 14B) via Ollama on a developer's laptop, with a layered architecture enabling easy extension into a full AI agent. The project emphasizes architectural best practices such as dependency injection, protocol-based interfaces, structured logging, and data validation with Pydantic. Phase 1 delivers a CLI chatbot that maintains conversation context but does not yet integrate real data or tools. The architecture supports future phases adding tool calls, retrieval, and state machines to evolve into a functional agent. The project is a portfolio/educational effort, not a deployed product.
Use Case
Opening the operator briefing
Pulling the full operator breakdown, tooling context, and verification notes.