Dremio Special Edition · Published by Wiley
The plain-English guide to autonomous AI systems
Everything data leaders, engineers, and analytics teams need to understand agentic AI — from how LLMs work to deploying agents on your data infrastructure without a PhD.
Each chapter stands on its own — dive into the topic most relevant to you, or read cover to cover. Either way, you'll come away with a clear mental model of how agentic AI works and how to put it to work.
Introducing Agentic AI
What makes a system truly agentic — autonomy, planning, tool use, and context persistence. Plus how AI agents differ from standard LLMs.
Understanding Text with LLMs
Tokenization, embeddings, transformers, and context windows on how LLMs actually process and generate language under the hood.
Boosting LLM Performance
Fine-tuning vs. RAG vs. prompt engineering and when to use each method and how to combine them for best results in enterprise environments.
Getting to Know MCP
How the Model Context Protocol became the open standard for connecting LLMs to external systems — databases, APIs, and your data lakehouse.
Resources, Tools & Prompts
A deep dive into the three things MCP servers exchange with LLMs and how sampling enables multi-step agent reasoning workflows.
Ten Benefits & Risks
Autonomous problem solving and reduced cognitive load on one side. Cascading errors, security gaps, and bias amplification on the other.
This isn't a whitepaper. It's a practical, approachable guide covering every concept you need from the foundational technology to deployment strategy.
LLM vs. AI Agent
| Capability | Standard LLM | AI Agent |
|---|---|---|
| Input Type | Prompt only | Prompt + Tools + Memory |
| Memory | Context window | External memory |
| Reasoning | Single-shot | Multi-step |
| Action-taking | ✕ No | ✓ Yes |
| Autonomy | None | Goal-oriented |
| Adaptability | Static behavior | Dynamic |
From the C-Suite to the data engineering team and if you're making decisions about where and how to use AI in your organization, this book is for you.
CDOs, VPs of Analytics, and data strategy leads building the AI roadmap
Building the AI roadmap and evaluating where agentic systems fit into the broader data strategy.
Engineers designing the infrastructure that agentic AI will run on
Designing the infrastructure that agentic AI will run on — lakehouses, pipelines, and data access layers.
Analysts and data scientists who will work alongside AI agents day-to-day
The practitioners who will work alongside AI agents day-to-day and need to understand what they can and can't do.
Executives evaluating agentic AI investments and organizational impact
Evaluating agentic AI investments and understanding the organisational impact before committing budget and resources.
No cost, no commitment. Just practical knowledge to help your team navigate the agentic AI era with confidence.
Fill out the form below for instant access to your free copy of Agentic AI For Dummies.
William Martin, PhD, is EMEA Evangelist at Dremio, where he brings over 15 years of data industry experience spanning roles at CERN, Deloitte, and Tamr. Beginning his career in statistical analysis for particle physics at the Large Hadron Collider, William has since worked across diverse industries including banking, logistics, healthcare, and defence. As a founding member of Tamr's EMEA branch and later their principal technical engineer for APAC expansion, he specialised in entity resolution and data quality solutions for Global 2000 companies. Now at Dremio, William travels across Europe speaking at meetups and conferences about Apache Iceberg, data lakehouses, and open-source analytics.
