Leanpub Header

Skip to main content
Go to Leanpub.com

The Field Deployed Engineer's Handbook

Production AI in the Real World

The Field Deployed Engineer's Handbook
This book is 90% completeLast updated on 2026-10-12
+15% in the last 30 days
Daniel Vaughan

Build production AI alongside your customers, from the first conversation to a system their team can run. Learn the engineering, judgment, and people skills of forward-deployed work in a practical handbook you can use across vendors and platforms.

Minimum price

$14.99

$24.99

You pay

Author earns

$

Also available for 1 book credit with a Reader Membership

PDF
EPUB
WEB
APP
About

About

About the Book

Build AI that works in a customer’s world, and leave their team ready to take it forward.

That is the work of a forward-deployed engineer. It means understanding how a business operates, connecting AI to existing systems, deciding what an agent should be allowed to do, and earning the trust of the people who will depend on it.

The Field Deployed Engineer’s Handbook takes you through the full engagement, from the first customer conversation to production, adoption and handover. It combines engineering principles you can apply across platforms with the judgement and people skills the work demands.

You will learn how to:

  • Identify problems worth solving, uncover hidden constraints and agree what success looks like.
  • Turn an ambiguous request into an architecture you can explain and defend.
  • Build agentic systems using retrieval, context, tools and orchestration.
  • Introduce coding agents into software teams with clear boundaries and effective review.
  • Evaluate system behaviour, control access and plan for failures in production.
  • Earn stakeholder trust, measure business value and help people adopt new ways of working.
  • Hand over a system the customer’s team can operate, maintain and extend.

You will also learn when to simplify a design, challenge a requirement or question whether AI is the right solution. For those building an FDE practice, the book explores how to turn lessons from individual engagements into reusable knowledge and better delivery across the team.

For engineers and architects moving into FDE work, practitioners already embedded with customers, and leaders building an FDE team. The book assumes professional software engineering experience.

Share this book

Categories

Agentic Engineering

Feedback

Contact the author

Author

About the Author

Daniel Vaughan

Daniel Vaughan is a technology leader and software architect based in the United Kingdom, specialising in agentic AI. He has spent approaching thirty years across enterprise, startup, and academic settings, with a career-long focus on engineering quality and developer productivity. His work now centres on the shift where AI stops being an experiment and becomes core to how organisations build software.

Daniel is Head of Forward Deployed Engineering at HCLTech AI Labs, where he leads a global practice embedding engineers and coding agents inside complex enterprise environments to take production AI from prototype to live systems. He built the practice from the ground up, running lean regional pods in which a shared architect, a small number of engineers, and a team of agents working through tools such as Codex, Claude Code, Cursor, and GitHub Copilot deliver at the output of a much larger team.

Before HCLTech AI Labs, Daniel was director of software engineering at Mastercard in London, leading cloud strategy and architecture for real-time payment products in a highly regulated financial services environment. Earlier, he spent eight years at the European Bioinformatics Institute in Cambridge, moving from software engineer into engineering leadership, working on the same problems of software quality and developer productivity that he now solves with agentic tooling.

He is the author of Cloud Native Development with Google Cloud (O'Reilly, 2024) and Ext GWT 2.0: Beginner's Guide (Packt, 2010), a Google Developer Expert, and a Green Software Champion. He writes about agentic engineering and AI-assisted development at blog.danielvaughan.com.

The Leanpub Podcast

Podcast

Podcast Episode

Contents

Table of Contents

Preface: How This Book Is Written

  1. Two reference points
  2. Why the book avoids products and exams
  3. The shape of the book
  4. How each chapter works
  5. How to read it

The Role

Chapter 1: The Forward Deployed Engineer

  1. The Production Gap That Created a Profession
  2. What a Forward Deployed Engineer Actually Does
  3. From Palantir Curiosity to Industry Practice
  4. Who Hires FDEs, and the Four Archetypes
  5. What Is the Function For?
  6. Why AI Makes Field Engineering Essential
  7. FDE Versus Everything Else
  8. Real Engineering or Fancy Consulting?
  9. The Pod and the FDE
  10. Do You Need an FDE?
  11. The Harness Engineering Connection
  12. The Three Running Engagements
  13. The FDE Mindset
  14. Key Takeaways
  15. Field Exercise
  16. Scenario Questions

Chapter 2: The Shape of the Role

  1. Convergence
  2. The Common Technical Core
  3. Field Judgements: The Portable Rules
  4. Two Engineering Dimensions
  5. The Ownership Boundary
  6. The Lifecycle Arc
  7. What Can Be Assessed, and What Only Deployment Shows
  8. Knowledge Half-Life
  9. Engineering Restraint as a Graded Skill
  10. Key Takeaways
  11. Field Exercise
  12. Scenario Questions

Chapter 3: The Capability Model and the Common Skills Map

  1. The T-Shaped Field Engineer
  2. The Five Capability Families
  3. The Common Skills Map
  4. Judgement: The Dimension That Runs Through All Five
  5. The Maturity Model
  6. The Minimum Viable Stack, as Capabilities
  7. The Learning Path
  8. Key Takeaways
  9. Field Exercise
  10. Scenario Questions

Chapter 4: The Honest Career and Evidencing Your Capability

  1. Why Engineers Are Interested
  2. Will This Keep Me Technical?
  3. A Realistic Week
  4. Travel and Burnout
  5. Career Paths
  6. Reading the Advert
  7. The Interview
  8. Evidencing Your Capability
  9. Compensation
  10. Negotiating the Offer
  11. Staying Current Without Chasing Products
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

The Method

Chapter 5: Discovery in the Real World

  1. The Gravel Track
  2. The First Week Determines Everything
  3. Preparing for the First Week
  4. Three Lenses, Not Three Phases
  5. The Five-Day Discovery Protocol
  6. Stakeholder Interviews
  7. The Discovery Output
  8. Key Takeaways
  9. Field Exercise
  10. Scenario Questions

Chapter 6: Discovering the Work

  1. Two Processes
  2. The Prototype Is the Instrument
  3. The Process Is Already Written Down, in the ERP
  4. Choosing the Unit of Work
  5. Watching People Work
  6. The Process Specification
  7. Exception Archaeology
  8. Deciding What the Agent Does
  9. Measuring the Baseline Before You Change It
  10. Validating the Specification
  11. Discovery Continues After the Build
  12. Field Exercise
  13. Key Takeaways
  14. Scenario Questions

Chapter 7: Discovering What People Know

  1. Four Seconds
  2. Four Kinds of Knowledge
  3. Why Interviews Fail
  4. Elicitation
  5. Writing Knowledge Down
  6. The Data Half
  7. Bias, and What You Are Encoding
  8. What Not to Capture
  9. Knowing Whether It Worked
  10. Field Exercise
  11. Key Takeaways
  12. Scenario Questions

Chapter 8: Discovering the Systems, Data and Tools

  1. Eleven Applications
  2. Inventory From the Desk Outward
  3. The Tool Inventory
  4. The Interface Ladder
  5. From Tools to Actions
  6. Who Is the Agent?
  7. The Data: Structured, Semi-Structured and Unstructured
  8. Observability and Somewhere to Fail
  9. Feasibility, and the Honest Report
  10. Key Takeaways
  11. Field Exercise
  12. Scenario Questions

Chapter 9: From Outcome to Defensible Design

  1. The Translation Problem
  2. Writing Outcome Statements
  3. Start by Not Using AI
  4. The Design-Decision Framework
  5. From Outcome to Plan
  6. Choosing the Delivery Vehicle
  7. Demo-Driven Development Without Demo Theatre
  8. Managing Scope Without Losing Momentum
  9. Writing Plans That Both Sides Trust
  10. Worked: Engagements A and B From Outcome to First Design
  11. Key Takeaways
  12. Field Exercise
  13. Scenario Questions

Ground

Chapter 10: Models, Instructions and Output Contracts

  1. Choosing the Model
  2. Prompting, Retrieval, Long Context or Adaptation?
  3. Instructions That Work
  4. Output Contracts
  5. Worked: Engagement B’s Invoice Extraction Contract
  6. Key Takeaways
  7. Field Exercise
  8. Scenario Questions

Chapter 11: Context Engineering and Memory

  1. What Enters the Context, and Why Less Is More
  2. The Four Operations
  3. Trimming, Pinning and Positioning
  4. Compaction and Context Editing
  5. Context Failure Modes
  6. Memory: Session State and Durable Memory
  7. A Layered Strategy
  8. Worked: Engagement A’s Multi-Issue Conversation
  9. Key Takeaways
  10. Field Exercise
  11. Scenario Questions

Chapter 12: Retrieval, Grounding and Documents

  1. When Retrieval Helps and When It Distracts
  2. Ingestion: The Ceiling on Everything Else
  3. Chunking, Embedding and Search
  4. Pre-Loaded or Agentic Retrieval
  5. Permission-Aware Retrieval
  6. Honest Grounding
  7. Freshness
  8. Diagnosing Retrieval Stage by Stage
  9. Worked: Engagement B’s Regional Policy Assistant
  10. Key Takeaways
  11. Field Exercise
  12. Scenario Questions

Orchestrate

Chapter 13: The Agent Loop, and When Not to Use One

  1. Workflows and Agents
  2. The Four Primitives
  3. Writing the Loop Without a Framework
  4. Stop Conditions, Done Properly
  5. Hardening the Bare Loop
  6. Recognising the Loop Inside Every SDK
  7. Agent Patterns in Practice
  8. Worked: Engagement A’s Half-Sentence Answers
  9. Key Takeaways
  10. Field Exercise
  11. Scenario Questions

Chapter 14: Tools and Integration Contracts

  1. The Job Is Integration
  2. A Tool Is a Typed Contract
  3. Descriptions Are the Selection Interface
  4. Fewer, Sharper Tools
  5. Structured Errors
  6. Idempotency Before Retries
  7. Synchronous, Asynchronous and Streaming
  8. Enterprise Integration Patterns
  9. Secrets and Orchestration Stay Server-Side
  10. The Smallest Useful Vertical Slice
  11. Worked: Engagement A’s Refund Tool
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Chapter 15: Protocols, Delegation and Discovery

  1. Why Protocols
  2. Tool Protocols: The Design Ideas
  3. Securing and Scoping Servers
  4. Agent-to-Agent: Delegation, Not Calls
  5. Registries and Discovery
  6. The Protocol Server as a Handover Artefact
  7. Worked: The Bank’s Change-Management Server
  8. Key Takeaways
  9. Field Exercise
  10. Scenario Questions

Chapter 16: Multi-Agent Systems, Escalation and Handoffs

  1. When Multiple Agents Earn Their Cost
  2. The Five Primitives
  3. Coordinators and Subagents
  4. Error Propagation
  5. Escalation Design
  6. Structured Handoff to People
  7. Worked: Engagement A Escalates a Policy Exception
  8. Key Takeaways
  9. Field Exercise
  10. Scenario Questions

Chapter 17: Engineering with Coding Agents

  1. The Second Engineering Dimension
  2. The Harness, Applied
  3. Agent Readiness
  4. Configuration as Code
  5. Plan, Then Execute
  6. Sandboxes and Blast Radius
  7. Agents in Continuous Integration
  8. Migrations and Refactoring at Scale
  9. Measuring the Change
  10. What Changes for the Team
  11. The FDE’s Own Pipeline
  12. The Agentic-Era Anti-Patterns, Revisited
  13. Worked: Engagement C’s Migration Programme
  14. Key Takeaways
  15. Field Exercise
  16. Scenario Questions

Prove

Chapter 18: Evaluation and Release Gates

  1. The Model Is Not the Product
  2. Golden Sets
  3. What to Evaluate
  4. Choosing the Evaluator
  5. Testing Non-Determinism
  6. Variance or Regression?
  7. Release Gates
  8. The Ratchet Principle
  9. Eval-Driven Development
  10. Evaluations Are an Asset the Customer Owns
  11. Worked: Two Suites
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Chapter 19: Observability, Troubleshooting and Cost

  1. Seeing What It Does
  2. Signals That Lead to a Decision
  3. The Troubleshooting Playbook
  4. Drift, Loops and Tool Latency
  5. Cost and Latency Are Release Gates
  6. Protecting Downstream Systems
  7. Worked: Engagement B’s Latency Regression
  8. Key Takeaways
  9. Field Exercise
  10. Scenario Questions

Chapter 20: Identity, Security and Access for Agents

  1. Who Is the Agent?
  2. Delegated Authority
  3. Three Identity Patterns
  4. Least Privilege by Removal
  5. Secrets and Credentials
  6. The Threat Model
  7. Defences: Architectural First
  8. Sandboxing and Egress
  9. Personal Data and Privacy-Aware Logging
  10. The Security Conversation
  11. Worked: Two Designs
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Chapter 21: Safety, Governance and Human Oversight

  1. Defence in Depth
  2. Deterministic Enforcement
  3. Approval Tiers
  4. Risk-Calibrated Autonomy and Human Oversight
  5. Audit, Lineage and Decision Logs
  6. Regulation as Design Requirements
  7. Responsible AI in Practice
  8. Worked: Two Controls
  9. Key Takeaways
  10. Field Exercise
  11. Scenario Questions

Chapter 22: Shipping and Running AI Systems

  1. Why Demos Lie
  2. Production Readiness: Six Dimensions
  3. Choosing the Runtime
  4. LLMOps: Prompts, Configuration and Model Pins Are Release Artefacts
  5. What Must Be Tested on Every Change
  6. Release and Rollback
  7. Monitoring for People Who Did Not Write the Code
  8. The Environment Contract
  9. Incident Response for AI
  10. The API You Leave Behind
  11. Worked: Engagement C’s Configuration, From Pilot to Organisation-Wide
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

The Human Side

Chapter 23: Working with Customers

  1. The Order-Taker Trap
  2. The Trust Spectrum
  3. Explaining AI Limitations Without Undermining Confidence
  4. Handling Sceptics and Champions
  5. Negotiating Trade-Offs
  6. Making Invisible Progress Visible
  7. Field Survival
  8. The Permanent Crutch Anti-Pattern
  9. Working in the Triangle
  10. Key Takeaways
  11. Field Exercise
  12. Scenario Questions

Chapter 24: Judgement Under Uncertainty

  1. Why Judgement Is the Job
  2. Facts, Assumptions and Unknowns
  3. When to Trust Your Intuition
  4. Sizing the Decision: One-Way and Two-Way Doors
  5. Before You Commit: The Premortem
  6. Writing It Down: The Decision Memo and the Decision Record
  7. Defending a Design
  8. Recommending Less
  9. Saying No, and Saying “Not Yet”
  10. Escalating Outside Your Boundary
  11. Integrity Under Commercial Pressure
  12. How the Field Judgements Fit Together
  13. Worked: Three Decisions
  14. Key Takeaways
  15. Field Exercise
  16. Scenario Questions

Chapter 25: Value, Commercials and the Business Case

  1. Why the FDE Owns the Value Case
  2. The Value Case
  3. Total Cost of Ownership and Cost to Serve
  4. Token Economics as a Business Conversation
  5. Proofs of Concept That Test Value
  6. Measuring Workflow Impact
  7. Commercial Models and What They Demand
  8. Statements of Work for Probabilistic Systems
  9. Land and Expand Without Dependency
  10. When to Walk Away
  11. Worked: Two Value Cases
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Chapter 26: Adoption, Handover and Product Feedback

  1. The FDE’s Real Deliverable
  2. The Autonomy Ladder
  3. Training by Doing, Not by Telling
  4. The Agentic Pod
  5. Handover Milestones
  6. Documentation That Survives
  7. The Harness as Ultimate Handover Artefact
  8. Change Management at Scale
  9. The N-to-1-to-N Lifecycle
  10. Separating Customer-Specific from Reusable
  11. The Field Report
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Chapter 27: Leading and Scaling the Practice

  1. The View from Above
  2. The Consulting Trap
  3. Should You Build an FDE Function at All?
  4. The FDE Lead Role
  5. Staffing and Hiring
  6. Quality Standards for Field Deliverables
  7. Career Paths
  8. Portfolio Visibility
  9. Building Reusable Harness Components
  10. What Firm Designations Demand
  11. A Complete End-to-End Scenario: Project Atlas
  12. Key Takeaways
  13. Field Exercise
  14. Scenario Questions

Conclusion

  1. The harness is the job, and the gap is why the role exists
  2. Concepts outlast products
  3. Contracts are where you put the pressure
  4. Non-determinism is an engineering problem, not an excuse
  5. Judgement is what no assessment can certify
  6. What you leave behind
  7. Where to go next

Appendix A: Field Judgements Index

  1. Candidate judgements
  2. The five ideas underneath

Appendix B: Concept Glossary

Appendix C: Scenario Question Reasoning

  1. Chapter 1
  2. Chapter 2
  3. Chapter 3
  4. Chapter 4
  5. Chapter 5
  6. Chapter 6
  7. Chapter 7
  8. Chapter 8
  9. Chapter 9
  10. Chapter 10
  11. Chapter 11
  12. Chapter 12
  13. Chapter 13
  14. Chapter 14
  15. Chapter 15
  16. Chapter 16
  17. Chapter 17
  18. Chapter 18
  19. Chapter 19
  20. Chapter 20
  21. Chapter 21
  22. Chapter 22
  23. Chapter 23
  24. Chapter 24
  25. Chapter 25
  26. Chapter 26
  27. Chapter 27

Appendix D: Common Skills Map Self-Assessment

  1. Part I: The role
  2. Part II: The method
  3. Part III: Ground
  4. Part IV: Orchestrate
  5. Part V: Prove
  6. Part VI: The human side
  7. Reading your scores

The Leanpub 60 Day 100% Happiness Guarantee

Within 60 days of purchase you can get a 100% refund on any Leanpub purchase, in two clicks.

See full terms...

Earn $8 on a $10 Purchase, and $16 on a $20 Purchase

We pay 80% royalties on purchases of $7.99 or more, and 80% royalties minus a 50 cent flat fee on purchases between $0.99 and $7.98. You earn $8 on a $10 sale, and $16 on a $20 sale. So, if we sell 5000 non-refunded copies of your book for $20, you'll earn $80,000.

(Yes, some authors have already earned much more than that on Leanpub.)

Learn more about writing on Leanpub

Free Updates. DRM Free.

If you buy a Leanpub book, you get free updates for as long as the author updates the book! Many authors use Leanpub to publish their books in-progress, while they are writing them. All readers get free updates, regardless of when they bought the book or how much they paid (including free).

Most Leanpub books are available in PDF (for computers) and EPUB (for phones, tablets and Kindle). The formats that a book includes are shown at the top right corner of this page.

Finally, Leanpub books don't have any DRM copy-protection nonsense, so you can easily read them on any supported device.

Learn more about Leanpub's ebook formats and where to read them

Write and Publish on Leanpub

You can use Leanpub to easily write, publish and sell in-progress and completed ebooks and online courses!

Leanpub is a powerful platform for serious authors, combining a simple, elegant writing and publishing workflow with a store focused on selling in-progress ebooks.

Leanpub is a magical typewriter for authors: just write in plain text, and to publish your ebook, just click a button. (Or, if you are producing your ebook your own way, you can even upload your own PDF and/or EPUB files and then publish with one click!) It really is that easy.

Learn more about writing on Leanpub