I have spent fifteen years building software at the intersection of enterprise architecture and product engineering — at Migros, Scout24, and through my own consultancy, Squibble, based in Switzerland. Alongside the client work, I built obseed — a sports analytics platform that processes health and performance data from wearables and training logs. That work is why the problem Linubra solves felt personal: I had more structured data about my body than most people ever collect, and no qualitative context to explain what the numbers actually meant.
Why I built this
obseed could tell me my resting heart rate was elevated for eleven days in a row. It could not tell me that those eleven days followed a difficult client negotiation, two late nights debugging a production issue, and a missed training week. That context lived in voice memos, calendar notes, and my own head — scattered, unconnected, and gone by the time the next training block started.
The second-brain tools I tried were honest about their trade-off: they gave you a canvas and asked you to do the connecting. I resented spending cognitive energy on filing instead of thinking. I wanted something that built the graph while I lived my life, not after. Linubra exists because that tool did not.
What I know
I have been writing production Ruby on Rails since version 4. Linubra's backend is Rails 8 in API mode — the same stack I have used for a decade of client work at Squibble, because I trust it at scale and can reason about every layer without reaching for documentation. The data model is PostgreSQL 18 with pgvector: a deliberate choice over a dedicated graph database, because I wanted something I could run, inspect, and debug myself rather than hand off to a managed service I do not control.
The health data work at obseed gave me the domain knowledge behind Linubra's Knowledge Graph. Quantitative data — heart rate variability, sleep scores, training load — is only legible when paired with qualitative context. I have spent years building the infrastructure to collect the numbers; Linubra is the infrastructure to collect the story that makes the numbers interpretable.
I hold a BSc in Computer Science from the Zurich University of Applied Sciences (ZHAW) and have been building commercially-deployed software since 2008.
What Linubra is not
It is not a startup with a growth team optimising onboarding funnels. It is a tool I use every day, built by one engineer who is deeply allergic to the things he finds annoying in software. That makes the decisions fast and the opinions strong.
The philosophy page explains the design convictions in more depth. The journal is where the thinking happens in public.
Get in touch
Questions, feedback, or an interesting use case: hello@linubra.me. Or find me on LinkedIn.