Resources

Practical AI, explained by someone who deploys it.

Free articles on putting AI to work in a real business — no hype, no fear. Written by Dennis Savic, founder of D3PLOY, from work inside actual businesses with real staff, real customers and real consequences.

Foundations

8 articles

Start here: who this is for, what AI actually is, and the tools I use every day.

Nº 013 min read

Who Am I, and Why This Page Exists

A mechanical engineer who built electric motorcycles, learnt some hard lessons, and now spends his days putting AI to work inside real businesses.

  • about
  • story
Read article →
Nº 023 min read

What D3ploy Actually Does (and Why I'll Mention It Sometimes)

A short, honest explainer of the business behind this page: review the flow, deploy the software and AI, then partner to keep leveraging what comes next.

  • about
  • d3ploy
Read article →
Nº 033 min read

Should I Use AI? Yes. Here's Where to Start

The answer is one hundred percent yes. The real questions are where and how, and this is the exact process I use to answer them.

  • strategy
  • process-mapping
  • getting-started
Read article →
Nº 043 min read

Your Business Is a Funnel, Fix the Narrow Part First

Why pouring AI into lead generation before your operations can scale is the most expensive mistake in small business AI.

  • strategy
  • operations
  • funnel
Read article →
Nº 053 min read

What Is AI, Actually?

A plain language tour of the main types of AI you'll actually encounter: language models, vision, audio, generative media, and coding harnesses.

  • fundamentals
  • explainers
Read article →
Nº 063 min read

What Is an LLM? (And Why I Use Claude)

How large language models actually work, why they sometimes get things wrong, why models differ, and why Claude is my daily driver.

  • fundamentals
  • llm
  • claude
Read article →
Nº 073 min read

The Claude Models Explained: Haiku, Sonnet, Opus, and Fable

What the different Claude models are, how they differ, and exactly which one I reach for on which kind of work.

  • claude
  • models
  • tools
Read article →
Nº 083 min read

Phone, Chat, Web, or IDE: The Four Ways I Use Claude

Claude lives in more places than most people realise. Here is every way to access it, and how I split my own day across them.

  • claude
  • workflow
  • tools
  • how-to
Read article →

Fundamentals

8 articles

The core skills: prompting, projects, agents, when to trust the output, and what to keep out of AI.

Nº 093 min read

How to Prompt: Context Is Everything

The difference between mediocre and brilliant AI output is rarely the model. It is the briefing. Here is how to brief like a pro.

  • fundamentals
  • prompting
  • how-to
Read article →
Nº 102 min read

Projects: Give Claude a Standing Memory of Your Business

Stop re explaining your business in every chat. Projects let you brief Claude once and benefit every conversation after.

  • claude
  • projects
  • workflow
  • how-to
Read article →
Nº 113 min read

What Is an AI Agent? From Chatbot to Workforce

Chatbots answer, agents act. A plain language ladder from asking questions to delegating whole jobs, and where automation fits.

  • fundamentals
  • agents
  • explainers
Read article →
Nº 123 min read

Trust and Verify: Using AI Without Getting Burnt

AI is a brilliant colleague who occasionally states nonsense with total confidence. Here is a practical system for catching it.

  • fundamentals
  • accuracy
  • how-to
Read article →
Nº 133 min read

What Not to Paste into AI: Data Safety for Business Owners

A plain English guide to using AI heavily without leaking the things that matter. No paranoia, just a sensible system.

  • fundamentals
  • privacy
  • security
  • business
Read article →
Nº 143 min read

Research Mode: Hours of Homework While You Make Coffee

Claude's Research feature turns a vague question into a cited report by reading dozens of sources for you. Here is when and how to use it.

  • claude
  • research
  • how-to
  • productivity
Read article →
Nº 354 min read

What Is an API? (And When Should You Use One)

The plumbing that lets software talk to software, explained without jargon: why your business already runs on APIs, and the questions to ask before you buy your next tool.

  • fundamentals
  • api
  • integration
  • explainers
Read article →
Nº 364 min read

What Is an MCP? (And When Should You Use One)

The Model Context Protocol is how AI gets hands: a standard way to plug models into your business systems. What it is, how it relates to APIs, and when it earns its keep.

  • fundamentals
  • mcp
  • agents
  • integration
  • explainers
Read article →

Business

6 articles

The decisions: what AI really costs, why pilots die, build vs buy, and bringing your team along.

Nº 153 min read

How to Find Your Bottleneck: A Practical Worksheet

The funnel article told you why operations must scale first. This one shows you exactly how to find the narrow point, in about an hour.

  • strategy
  • operations
  • worksheet
  • bottleneck
Read article →
Nº 163 min read

The Real Cost of AI: Subscriptions, Tokens, and Honest ROI Maths

What AI actually costs a small business, including the hidden lines, and a back of the envelope ROI method anyone can run.

  • strategy
  • cost
  • roi
  • business
Read article →
Nº 173 min read

Why Most AI Pilots Die (and How to Run One That Lives)

The technology usually works. The pilot dies anyway. An autopsy of the common causes of death, and the fix for each.

  • strategy
  • adoption
  • change-management
  • business
Read article →
Nº 183 min read

Build, Buy, or Managed? An Honest Framework

Three ways to get AI working in your business, the real trade offs of each, and the questions that decide it, bias declared up front.

  • strategy
  • decision-framework
  • business
Read article →
Nº 193 min read

Getting Your Team on Board: The 'Will This Take My Job' Conversation

AI rollouts succeed or fail on trust. How to have the honest conversation, involve the right people, and make the tool a gift rather than a threat.

  • strategy
  • team
  • change-management
  • leadership
Read article →
Nº 203 min read

Where Humans Stay: The Work AI Shouldn't Do

The point of automating the repetitive work is to protect the human work. Here is what I deliberately keep human, and why it pays.

  • strategy
  • philosophy
  • business
Read article →

Use cases

5 articles

Concrete wins: phones answered, inboxes tamed, quotes in minutes, meetings into action lists.

Build in public

5 articles

Real builds documented as they happen, numbers included, warts and all.