Aaron Lysaght Walsh

Chicago, IL  /  BSc Computer Science with Data Science, University College Dublin, 2025

Where technology,
data and
people meet.

Hi, I'm Aaron. I studied Computer Science with Data Science at University College Dublin and moved to Chicago in September 2026.

I'm most interested in the space between technical work and the people who rely on it: turning data into something a leadership team can act on, and deciding where AI can and can't be trusted. I've also spent years in customer-facing roles, so I'm looking for work that uses both sides.

See my work Get in touch

Open to technology and AI consulting, business analytics, solutions and customer-facing technology roles in Chicago.

Selected work

An AI product I specified, a commercial analytics internship, and a controlled research study. Each one is about the same thing: technology people can trust and act on.

In development Independent AI project, 2026

GetMeAJob An AI job-application engine that won't make things up

Generic AI tools can write applications that sound convincing but invent experience. GetMeAJob is built around one rule: every claim must trace back to real evidence. It turns several resumes into one profile, researches each employer, and will say when a role isn't worth applying for.

I defined the product and its rules and directed the build. Claude designed the architecture and wrote the code.

My role
Product owner and designer
Built by
Claude (architecture, code and tests)
Status
Phase 1 prototype, running locally; not a production product
Read the case study View the code on GitHub
Fig. 1 Every claim traces back to evidence. The tailored resume's integrity panel shows an unsupported skill ("Salesforce") removed and an embellished number fixed automatically. Screenshots show a fictional applicant and simulated AI output. Any scores shown come from test data, not real-world results.

Internship Data Science Intern, Dublin, March–July 2024

Harris Group Turning automotive data into business insight

I built Power BI dashboards and analyzed automotive inventory, sales and vehicle movement data across regions, then presented the findings to the Managing Director, executives and department leaders.

Sales and inventory data was spread across several reporting sources, so I recommended consolidating them into a single star-schema data model. The recommendation was approved, the sources were merged, and the dashboards went on to be used in leadership reviews.

Read the case study

The business questions I worked on

  • Dead stock. Which automotive parts had no recorded sales for 180 days or more, and needed management attention?
  • Regional allocation. How did demand and stock compare across regions, including for electric and diesel vehicles?
  • Replenishment. What do historical usage and supplier lead times suggest for safety-stock levels?
  • Data foundations. How could fragmented reporting sources become one consistent model?

Original dashboards and company data are confidential and aren't shown.

The high-control video interface: a session timer and category filter above the player, an autoplay on/off toggle, a Next Video button, and a row of recommended videos with a labeled heading.
Fig. 5.2 The high-control interface participants used, with a session timer, category filter, autoplay toggle and labeled recommendations. Videos are third-party YouTube Shorts embedded for the study.

Completed Final-year research, University College Dublin, 2025

UCD Research Does giving people control change how they feel about a platform?

For my final-year research, I built two versions of a short-form video app. One gave people visible controls and explained why each video was recommended; the other played videos endlessly with no explanation.

  • 47 participants completed the study.
  • Participants using the high-control interface reported a significantly greater sense of agency.
  • No statistically significant difference in clicking behavior was detected between the two groups.

The takeaway for product teams: usage numbers alone can miss how people actually feel about a product.

Read the case study

What I bring

The mix of skills behind the work above.

Technical understanding
Python, SQL, Power BI and data modeling, with coursework in information security and artificial intelligence.
Analytical problem-solving
Starting from the business question, then finding out what the data can actually support.
Commercial awareness
Selling on commission, representing major brands to the public, and reading data in the context of real operating decisions.
Stakeholder communication
Explaining analysis to senior leaders and non-technical audiences in plain language.
AI and product curiosity
Deciding how an AI product should behave, what it should never do, and how its output can be checked.

About me

Not all of my experience happened at a keyboard. I've made sushi in a busy Dublin kitchen at YO! Sushi, sold ski holidays on commission, and spent four years at events and brand activations for companies including Google, Coca-Cola and Diageo.

Rugby has been part of my life for about 17 years, and I've lived, worked and traveled in several countries before settling in Chicago.

More about me

Let's talk

I'm based in Chicago and interested in technology and AI consulting, analytics, solutions and customer-facing technical roles.