I built an HR app. On a weekend. With no engineering background.
Andy Maren — March 25, 2026
I want to tell you about the strangest, most humbling, and genuinely exciting thing I've done this year.
Over the last several weeks, I built four functional HR applications. Not prototypes in the "I made a mockup in Figma" sense. Actual working software that runs on a phone. Real interfaces, real data flows, real logic.
I am not a developer. I have never written production code in my life. My background is HR and payroll implementation — Workday, mostly, which means I know configuration, business process logic, and data modeling, but my programming "skills" peaked at writing calculated fields in report builder.
What changed is a tool called Claude Code, combined with VS Code, a mobile development framework called Expo Go, and a UI component library called Gluestack. I've been building with these tools on weekends and evenings, and I want to share what I built and what I actually learned — because I think it has real implications for how HR leaders should be thinking about their own relationship with technology.
What I built
A custom learning management system integrated with Wistia
My first project. A lightweight LMS that hooks into Wistia — a video hosting platform — so that training content can be organized, assigned, and tracked without a bloated enterprise system in the middle. The idea came from conversations with clients who were paying significant amounts for LMS platforms they were using at about 20% capacity.
Building it forced me to think through what a learning system actually needs to do at its core: content delivery, progress tracking, completion reporting, and basic assignment logic. Stripping away everything else clarified why so many enterprise LMS products feel like overkill for what most mid-market organizations actually need.
A rewards and recognition app that pays workers in Bitcoin micro-transactions
This one raised some eyebrows when I described it. The concept: peer-to-peer recognition that actually carries a small monetary reward, settled in Bitcoin, delivered instantly to an employee's wallet. No gift cards, no "points" that expire in a catalog, no HR admin bottleneck.
I'm not suggesting every organization should do this — the regulatory, tax, and policy questions are real. But building it taught me something important about the gap between what digital currency makes technically possible in HR and what the industry has actually built for. That gap is large and it's closing fast.
A document transformation GPT for HR content
This one was built specifically for a consulting use case I kept running into: clients with enormous libraries of user guides, help documentation, and training materials in Workday — all formatted inconsistently, often outdated, frequently hard to find.
What I built was a tool that ingests that documentation, rewrites it to a consistent structure and format, and then runs a check against each new Workday release to flag content that may need updating. The time this could save an HR operations team is not small.
A natural language headcount reporting interface
The one that surprised me most in terms of what it revealed. A simple interface that lets an HR leader ask questions of their headcount data in plain English — "How many open roles do we have in operations that have been unfilled for more than 60 days?" "What's our voluntary attrition rate in the midwest region year to date?" — and get answers without opening a BI tool or submitting a request to analytics.
This one made me feel something I wasn't expecting: a little bit sad. Not because the technology is bad. Because I thought about how many hours I've watched HR business partners spend waiting for data that should be at their fingertips, and how much of that waiting is now technically unnecessary.
What I actually learned
The tools are genuinely different now. I don't want to overclaim this. Building these apps still required patience, iteration, and a willingness to read error messages at 11pm. But the gap between "I have an idea" and "I have a working prototype" is genuinely measured in days now, not months. That has not been true before.
My HR domain knowledge was the asset. The apps I built weren't technically sophisticated — any junior developer would look at them and find things to improve. What made them useful was that they were designed by someone who understood the actual HR workflow, the real pain point, the compliance consideration that needs to be built in from the start. That's not something Claude Code can supply. It's what I brought.
I now understand the vendor conversations differently. Having built something — even something simple — gives you a completely different frame for evaluating what vendors are selling. You understand what's actually hard and what's just positioned as hard. That's worth something.
The skills question for HR is real. I came into this with more technical comfort than most HR professionals because of my Workday background. And I still found parts of this genuinely hard. The HR professionals who are going to thrive in the next five years are the ones developing a working relationship with these tools now — not necessarily building apps, but understanding what's possible, staying curious, and refusing to treat "I'm not a tech person" as a permanent identity.
I'm going to keep building. I have a list of ideas — most of them are things I wished existed when I was in the chair. I'll share more as they come together.
If you're an HR leader who's been curious about this but hasn't started — start. Pick one small pain point. Describe it clearly. See what happens.
Update your curiosity.
— Andy
Andy Maren is the founder of Trailhouse Solutions, an advisory firm for HR and payroll leaders navigating the shift to agentic AI and modern HR technology. She has been building HR software prototypes using Claude Code, VS Code, Expo Go, and Gluestack UI.
