Chatbot: Add notebook for DIS2025 workshop

Author: wierdvanderhaar

topic/chatbot/table-augmented-generation/workshop/README.md

@@ -0,0 +1,68 @@
+# Timeseries QA with LLMs
+
+*Combine Telemetry Data and Manuals for Smarter Diagnostics.*
+
+---
+
+## Summary
+
+In this hands-on workshop, participants will build an AI-powered assistant that can answer questions using a combination of **timeseries telemetry data** and **structured machine manuals**. 
+
+Using **CrateDB** as a scalable time-series database and **OpenAI's large language models**, attendees will walk through how to:
+- Simulate telemetry data
+- Store and query with SQL
+- Convert natural language into structured SQL queries
+- Integrate external context (like manuals)
+- Generate explanations and visualizations based on live data
+
+This workshop combines **Table-Augmented Generation (TAG)** with best practices in prompt design, data enrichment, and telemetry monitoring — creating an interactive diagnostics assistant from scratch.
+
+By the end of the session, participants will have a working QA system capable of answering both technical and operational questions in natural language, using real telemetry data and contextual documentation.
+
+---
+
+## Key Takeaways
+
+- Understand how to build a **TAG (Table-Augmented Generation)** assistant using structured data and manuals
+- Learn how to simulate, store, and analyze **timeseries data** in **CrateDB**
+- Master prompt engineering techniques to guide LLMs in generating accurate SQL queries
+- Combine telemetry results with structured documentation (manuals) to generate full-context answers
+- Create **visualizations** from natural-language queries using **Matplotlib and Plotly**
+- Explore how LLMs can reason across structured (CrateDB) and unstructured (manuals) sources
+
+---
+
+## Requirements from Attendees
+
+- **Laptop + Charger** (for running Jupyter notebooks)
+- **Python Basics** (e.g. working with pandas, functions, etc.)
+- **Basic SQL Knowledge** (SELECT, WHERE, GROUP BY…)
+- **OpenAI API Key** (required for accessing GPT)
+
+---
+
+## Maturity Level
+
+**Intermediate**  
+This workshop is ideal for developers, data engineers, solution architects, and tech leads who have some experience with Python and data querying. No deep ML background is needed — the focus is practical application.
+
+---
+
+## Type of Workshop
+
+**Hands-On Lab**  
+Participants will interactively build, run, and modify a full-stack LLM-powered assistant. Exercises are incremental, exploratory, and designed for real-world relevance.
+
+---
+
+## Try It Out
+
+You can explore the full interactive notebook in two ways:
+
+### ▶️ Run It Directly in Google Colab
+Click below to launch the notebook in a ready-to-use Colab environment (no setup needed):
+
+[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/crate/cratedb-examples/blob/main/topic/chatbot/table-augmented-generation/workshop/telemetry-diagnostics-assistant.ipynb)
+
+### 📓 View the Notebook on GitHub
+[🧠 Timeseries QA Notebook (Jupyter)](https://github.com/crate/cratedb-examples/blob/main/topic/chatbot/table-augmented-generation/workshop/telemetry-diagnostics-assistant.ipynb)