Skip to content

Blog

What I tell clients, written down. 

The same questions come up on almost every first call. Do we need a custom model? What should we automate first? These are my answers, kept short.

All posts

4 posts, newest first

  1. A single glass sphere beside a structure assembled from connected glass blocks

    1 min read

    AI engineer vs ML engineer: which one does your business need?

    Clients often ask if I will train them a custom model. Almost never. Here is the difference between the two roles, and why it matters for your budget.

    • AI engineering
    • For founders
  2. A winding line of connected glass nodes with one node lit bright blue

    1 min read

    You don't have an AI problem. You have a workflow problem.

    When founders ask about AI they expect something exotic. Then we look at how the work flows, and the fastest wins turn out to be ordinary.

    • Automation
    • For business owners
  3. A layered glass structure with a glowing core built into its centre

    1 min read

    Why most AI features fail: bolted on vs architected

    A chatbot stapled to the corner of an app demos well and then quietly dies. The difference is whether the intelligence is part of the system.

    • Product
    • AI engineering