Under the Hood

What I’m building.

I design and build data, cloud, and AI systems at enterprise scale. The independent work is the other side — the same problems, from first principles, under real constraints, to pressure-test what I advise. About Samuel →

Index — three live projects

AskSam

Live · Independent backend

A grounded portfolio assistant extracted into an independently deployable service without moving its evidence or cloud resources.

AskSam · See grounding and boundaries in practice

A narrated, provider-free demonstration using the real interface and reviewed synthetic responses: AskSam grounds answers in public sources, returns its deterministic public-contact answer, and shows its explicit no-answer boundary.
Transcript

AskSam is a grounded portfolio assistant. It answers from reviewed public evidence, shows its sources, and knows when to stop.

Ask how Ebook to Audio works. The interface follows the real request flow, retrieves the relevant project notes, and returns a concise answer with its source still attached.

The same evidence model handles broader questions about leadership. The answer connects Samuel's working principles to the published values behind them.

For direct contact, AskSam uses a closed deterministic path. It returns Samuel's published LinkedIn profile without spending a model call or inventing another channel.

Ask for a private email address, and the boundary is explicit: no published evidence, no answer. Behind the simple interface, retrieval, safety, cost controls, and citation projection keep every response useful and bounded.

  • RAG
  • Grounded AI
  • AI Safety
  • AI FinOps
View project →Visit live ↗

nottoobad

Live

Trustworthy purchase decision support built around evidence, trade-offs, and an inspectable verdict.

nottoobad · See a purchase decision unfold

An editorial walkthrough built from first-party captures: nottoobad tests the buyer's routine, keeps the dossier inspectable, compares a simpler alternative, and reaches a structured Skip verdict.
  • Decision Intelligence
  • AI Governance
  • Grounded AI
View project →Visit live ↗

Ebook to Audio

Live

A document-to-audio system that preserves structure through recoverable AI processing.

Ebook to Audio · See a book become a navigable audiobook

An editorial walkthrough built from first-party captures: source analysis, voice selection, visible conversion progress, chapter navigation, and an answer grounded in the book.
  • Document Intelligence
  • Speech Synthesis
  • Resilient Pipelines
View project →Visit live ↗
The professional work surfaces the questions. Building independently is how I know the answers hold.
How I think and lead →

About

I align data, cloud engineering, and applied AI with business direction, from strategy through delivery and adoption.

Operating context

Most of my career has been spent where data systems meet regulated industries under commercial pressure. Close to a decade in financial services — building customer-experience measurement, audience identity, and marketing data infrastructure across banks at different scales and maturity levels. Then three years in luxury and fashion retail, directing analytics and applied AI at the scale of Zara, NET‑A‑PORTER, and Mytheresa. Since early 2026, directing cloud and applied AI architecture across EMEA at Microsoft.

How I lead

I separate problems from people and make priorities and trade‑offs explicit when engineering, product, and business needs compete. I connect executive direction with architecture and implementation across data, cloud, and AI. I document decisions and reusable practices so teams and systems can keep improving beyond one project.