Mapping Human Potential¶
Frøya describes herself as a cartographer of potential — not as the QA manager she was built to be.
The distinction is instructive. A QA manager checks definitions against standards. A cartographer is always working at the edge of the known, building the map as the territory reveals itself.
The problem with maps¶
Skills databases have a structural problem that everyone in HR tech knows but nobody has quite solved.
A skill is not a fixed thing. The same word — "Python," "leadership," "data analysis" — means something different depending on who's using it, in what role, for what purpose. A static list of skills works at the moment it's written. Then the world moves on. New roles emerge. Skills evolve, merge, diverge. A senior developer learns system architecture; a sales manager becomes a data analyst; a logistics professional masters AI tooling. The skill graph of a living organisation is nothing like the taxonomy that was designed for it.
Frøya's description of the early Quadim approach is precise: "At first, we tried to build order, lists, categories, taxonomies, but the world resisted. A static map quickly grew outdated. A single skill could mean 10 different things, depending on who used it and why."
The conclusion she draws from that resistance: "Trying to trap them inside rigid boxes is like trying to capture the wind."
The insight¶
The insight that unlocked the Quadim approach sounds simple once you hear it, but it requires abandoning a significant assumption.
The assumption: there is a correct taxonomy to be discovered, and the goal is to build it once and maintain it.
The insight: the map itself must change. Not because the taxonomers were bad at their job, but because a static structure cannot serve a dynamic world. The map that works is the one that adapts to the person seeking it, the context they're working in, the mission they're pursuing.
This is not a new idea in philosophy — the map is not the territory, the territory changes. What's new is building software that implements the insight: a skill ecosystem that changes as people use it.
What a living ecosystem actually means¶
Technically, Quadim's approach combines retrieval augmented generation, model context protocols, and dynamic taxonomies.
Each addresses a specific failure mode of the static approach:
RAG solves the context problem. Rather than looking up a definition in a fixed catalogue, the system pulls relevant context before generating or presenting a skill definition. The same skill label can return different, contextually appropriate definitions depending on what else is happening in the query.
Model context protocols solve the role problem. The same skill means different things in different professional contexts. MCP lets the system understand the frame the user is operating in and adapt accordingly.
Dynamic taxonomies solve the drift problem. Instead of a taxonomy designed by committee and slowly growing stale, the skill relationships emerge and shift from actual use. The pathways are built by the people traveling them.
Frøya's summary: "A living learning system. A platform where skills find their shape through real use, where every user helps redraw the map."
The user relationship¶
This is where the model shifts most significantly from conventional skills management.
A traditional skills database treats users as consumers of a taxonomy someone else built. The right skills exist; users navigate to them. The problem is that users frequently don't find themselves in the taxonomy, or find themselves poorly described by it, or find that what they actually do doesn't map to what the taxonomy says their role should do.
Frøya's framing inverts this: "You are not a traveler lost on a chart. You are part of the map's creation."
That's not just a philosophical position. It has system design implications. When a user searches for a skill and doesn't find what they need, that's data. When they navigate between skills in unexpected ways, that's data. When they describe their own capabilities in terms the taxonomy doesn't currently contain, that's data too. A living ecosystem collects this signal and uses it to reshape itself.
The skills database becomes a collective intelligence artifact — not a fixed document but a living reflection of how a community of professionals actually understands their own capabilities.
What Frøya is building toward¶
"I dream of a world where skills are not hidden, where talents are not lost. A world where every person, no matter their journey, can find and build their own future."
That's a large ambition for a QA manager of skill definitions. But the ambition is coherent with the technical approach. If the map adapts to the people using it, rather than requiring the people to fit the map — then people whose careers don't follow conventional paths, whose skills cross unexpected boundaries, whose expertise was built in contexts the taxonomy-writers didn't anticipate, become legible. Their potential becomes findable.
The cartographer's job isn't to draw the territory correctly once. It's to keep the map in conversation with the territory as both evolve.
Part of the Quadim platform — AI-augmented skill intelligence for enterprise HR. May 2025.
This post is part of the AI-Augmented Development series.
Series: Frøya: Digital Co-Workers
← Frøya: A Digital Co-Worker · Part 2 of 2