I already run Ollama locally and now need a Python-based chatbot that lives on my own server, talks fluently with the Microsoft Teams API, and uses MongoDB as its memory. The bot must do three core jobs: • answer user queries drawn from our internal data • manage simple scheduling requests inside Teams • push proactive notifications when certain data triggers fire Everything has to work completely offline from any external LLM service; the model must load through Ollama and stay inside our environment. A clean, well-commented Python codebase, plus set-up scripts and README, will be the expected hand-off. Please include straightforward instructions for connecting to the Teams endpoint and seeding/structuring the MongoDB collections so I can test end-to-end the moment I pull the repo.