From building websites to shaping products

My name is Praveen Mahato, but everyone calls me Ryan. I started as a developer, building online shops, widgets and interactive modules in Kathmandu, and leading my first team there. Every client’s business was different, so every site had to be shaped around the people it served. I didn’t know it yet, but that was product work.

Then I moved to the United States and worked on-site for Energy Dynamics, EF Hutton and Hubbard Broadcasting, and saw what happens at scale. At Hubbard my main job was analysing millions of user interactions: what people read, where they dropped off and what kept them coming back. The placement of an article or an ad changes how people read. Search terms tell you what people want before they tell you. Speed is a feature. I started to spend as much time in the analytics as in the code.

Then came six years at Cybersmarties in Limerick, shaping a social and monitoring platform for school children. I studied how students and teachers used it, worked with management on direction, and showed them what each user cost to serve before they decided how to grow. By the end, I was doing a product manager’s job with an engineer’s toolkit.

Today I run horsesales.ie as its owner, which means the product, the build, the marketing and the revenue are all mine. Alongside it, I’ve built AI-powered analysis products for financial data as a consultant. Both have sharpened the same belief: understand what a product is really for, shape it around the people using it, and keep improving it in small, measured steps.

Where AI fits

AI has become part of almost everything I do. As a consultant I built AI-powered analysis products for financial data, which meant researching model approaches, building a training corpus and fine-tuning a specialised model that was efficient to run. I design the algorithms and pattern-recognition logic that find signals in data, build RAG applications that answer from a company’s own content, run local LLMs, and experiment with generative video, audio and 3D on my own GPU workstation. Every day I work alongside Claude Code, Cursor and Copilot.

What I bring to AI is the same thing I bring to any product: a clear idea of the job it should do for real people, a way to measure whether it does it, and an eye on what it costs to run. More on how I build AI products.

Why the developer years matter

They are what set me apart. I can estimate effort honestly, spot technical risk early and talk to engineers as a colleague rather than a client. When it’s faster, I build the first version myself. It means less time between learning something about users and putting a better product in front of them.