Project Title: "Beyond Chatbots: Building a Generative AI Agent for Proactive Decision Support"
Abstract:
This group will create an advanced AI agent based on Retrieval-Augmented Generation (RAG) and generative AI. The AI agent will assist users not only with queries but will also proactively help generate new KPIs, dashboards, and reports based on real-time data and system context.
Repository Structure:
./
├── 📂 rag_service
│ ├── 📄 config.py
│ ├── 📄 constants.py
│ ├── 📄 rag_lib.py
│ └── 📄 utility.py
├── 📄 data.ipynb
├── 📄 milestone1.ipynb
├── 📄 milestone2.ipynb
├── 📄 milestone3.ipynb
├── 📄 rag_notebook.ipynb
└── 📄 requirements.txt
In the folder rag_service you can find all the scripts for the final delivery: rag_lib is the core file where the model is uploaded, the embeddings are created, and the prompt is chosen based on the query. The available prompts can be found in constants.py.
The final notebook is rag_notebook.ipynb, where you can see the mocked GUI interaction with our RAG system (i.e. the function rag_interaction contained in rag_lib.py).
The notebook data.ipynb was used for the first development of the system to mock the knowledge base from the available data provided by professors.