About

One person, one timezone.

I make operational data trustworthy. I audit systems against the client’s own source files, fix what doesn’t match, and add controls so bad data can’t get back in.

Available now · freelance · remote

Andrew Ryan

I’m Andrew Ryan, a freelance data and automation consultant based in Belgium. I’ve spent five years across analytics, operations automation and software delivery, most recently for international and humanitarian organisations as an external consultant with Relief Applications, a technology consultancy.

I build with Python, SQL/PostgreSQL and n8n. I use AI-assisted development (Claude Code) as a tool under the same rules as any other: every change is read, tested and reviewed, and then checked against real data before I call it done. Tests and code review say the code does what it was written to do; only the client’s own files say whether that was right.

In a team I work the usual way: branches and pull requests on GitHub, code review, and tests alongside each fix. At Relief Applications I work inside a codebase the team built.

I work directly with the people who use what I build, in English or French. You deal with me from the first call to the handover, with no account manager in between. I’m registered in Belgium as a sole trader, so the contract is with me directly.

Download my CV (PDF)  ·  LinkedIn

Experience

Where I’ve worked.

  1. Data & AI Consultant, external contractor · Relief Applications, technology consultancy for international and humanitarian organisations

    Audit, fixes and controls on a payroll allocation platform for an international human rights NGO. AI usage policy for a humanitarian accreditation body. AI readiness assessments for international institutions.

  2. Data & Automation Consultant · Andrew Ryan Consulting, independent practice, Belgium

    Supplier scraping and enrichment pipeline for Ten Lives, a regenerative community festival in Portugal, and its volunteer onboarding pipeline (in progress). Seven operational dashboard demos on synthetic data.

  3. Growth & Automation Advisor, contract · Rapidos.ai

    Automated knowledge bases and lead-qualification flows. Business requirements turned into n8n workflows and internal AI agents.

  4. Co-Founder · Galene Pathways, marine technology

    Internal data workflows and scripts for environmental data pipelines. User discovery to validate the data assumptions and the human-in-the-loop logic.

  5. Teaching Assistant, Data Science · Le Wagon

    Supported students through Python, SQL and machine-learning coursework: debugging, code review and one-to-one explanation.

  6. Operations Automation Specialist · Shippr, logistics

    Designed and maintained operational workflows for account access and documentation, kept consistent and compliant between operations and engineering.

  7. BI & Analytics · Havas Media

    Dashboards analysing user behaviour. Research on AI trends in marketing and personalisation, reported to the CEO.

Education
Data Science Bootcamp, Le Wagon, 2022: Python, machine learning, SQL, deep learning, Docker.
Languages
English and French: C2 (fluent). Dutch and German: B1. Spanish and Portuguese: A1.
Based
Belgium. Remote, Central European hours (CET/CEST).

How I work

Four steps, every time.

  1. Map the process

    How the work really flows, which files it runs on and where it breaks. The org chart and the process document are a starting point, not the answer.

  2. Build the smallest useful thing

    One problem solved end to end, in use early, so the next decision is made on something real rather than on a plan.

  3. Check it against real data

    The running system against your own files. Every difference is explained or fixed before anything is called done.

  4. Hand it over, documented

    Code, credentials and a written handover, so nothing depends on me being around.

What you can expect

  • One person from the first call to the handover.
  • Short written updates, in English or French.
  • Normal team practice: pull requests, code review and tests.
  • Work done in your environment, with your credentials.
  • A clear scope, agreed in writing before the work starts.
  • The code and the documentation are yours at the end.

What I won’t do

  • Name a client, or show their data, without their say-so.
  • Claim a number that nobody measured.
  • Call something done before it has been checked against real data.
  • Sell you a build when a process change would do.
  • Keep access to your systems after the handover.

Talk to me, not a sales team.

A 30-minute call about one problem, free. Hiring for a contract? The recruiters page has the short version.

Book a call → Write to me instead →