About

I'm Anthony. I run Atomic Scale, an AI studio focused on one thing: turning messy, manual business processes into reliable AI systems that people actually use.

I've spent the last 8+ years streamlining company processes with automation and applied AI. Not in theory. In real organizations, with real data, real constraints, and real resistance to change.

My work usually starts the same way: someone wants "AI".

What they actually need is clarity, prioritization, and a system that doesn't fall apart after the demo.

That's where I come in.

What I've done

Over the years, I've worked on 50+ AI and automation deployments, covering workflows like:

  • HR operations
  • Procurement and purchasing
  • Operations and back-office processes
  • Legal and compliance-adjacent use cases

Most of these projects were delivered for large companies, mainly in the UAE, where expectations are high and tolerance for half-working systems is low.

Before Atomic Scale, I was Account Executive and AI Strategy Lead at Aleria AI, where my job was to help companies figure out what to build, why it mattered, and how to actually get it into production.

Before that, I spent a few years in New York, working as a senior consultant for top companies in the US. That's where I learned a valuable lesson early on:

Slides don't fix processes. Systems do.

This mix business, strategy, and hands-on delivery is what shaped how Atomic Scale works today.

My point of view

After dozens of AI and automation projects, one thing is clear:

Most AI failures have nothing to do with the technology.

They fail because of how decisions are made.

Why AI initiatives usually fail

Here's what I see again and again:

  • The wrong problem is chosen too early

    Teams start with tools or trends instead of identifying where time, cost, or risk is actually concentrated.

  • "Let's build our own AI platform" is underestimated

    Building a custom AI platform sounds simple until you hit reality: understanding the business process, choosing the right architecture, integrating tools, and making the whole thing scalable and maintainable. Most teams underestimate this massively.

  • Data is treated as an afterthought

    Garbage in, garbage out. If the data isn't clean, structured, and focused on the right signal, the AI will never be reliable — no matter how good the model is.

  • Systems are built that nobody can operate

    If teams don't understand or trust the system, it won't last. Period.

What actually makes AI work

Successful AI initiatives look very different:

  • They start with one clearly defined, high-impact workflow

    Not "AI everywhere". One place where automation genuinely changes outcomes.

  • They understand the data first

    What exists, what's usable, what needs structure, and what actually matters.

  • They map the full process before building anything

    Flows, edge cases, responsibilities, and failure modes are clear upfront.

  • They're designed as systems, not experiments

    Reliability, monitoring, and human control are built in from day one.

  • They're built for autonomy

    Teams are trained to operate, maintain, and evolve the system without being dependent on me.

How this shapes Atomic Scale

I work on a small number of high-impact engagements at a time, own delivery end to end, and build systems that clients can understand, operate, and extend themselves.

The goal isn't to deploy more AI.

The goal is to deploy the right AI — and make sure it keeps working once I'm gone.

Ready to make one workflow disappear?

Let’s take one workflow and decide if it’s worth automating.