How My Rank is Calculated
The rank you see on this page is not a label I chose. It is the output of a calibration engine that reads every project I have logged and produces a score based on how the work is structured, not how long it took.
Here is how it works.
The Problem With Years of Experience
Years of experience tells you how long someone showed up. It says nothing about what they built, how deeply they practiced each domain, whether they kept using skills after learning them, or whether their output grew over time.
Two people can both have three years of experience. One shipped production systems across four domains and kept revisiting the same skills at increasing depth. The other completed courses and did small experiments that were never used again. They are not equivalent, but a YoE metric treats them as identical.
This engine tries to fix that.
Three Layers of Measurement
Layer 1: Sensors
The engine reads my project registry and computes 32 sensors from the raw data. These are concrete measurements like:
- How many skills I have actually used across real projects (not just listed)
- What fraction of those skills I kept using after the first time, versus abandoned
- How long my projects were on average, and what the longest one was
- How many distinct domains each project crossed
- How consistently I have been active across calendar years
- How recently I last shipped something
- Whether my output has been growing or plateauing over time
- The ratio of shipped projects to education and research
Each sensor is weighted by context. A deployed product counts more than a course. A published research paper counts more than a personal experiment. Education counts the least, because reading about something is not the same as building with it.
Layer 2: Career DNA
The 32 sensors are then compressed into an 18-number vector that describes the shape of the career. Think of it as a fingerprint. Two models build it.
The first model reads meaning. It turns the sensors into 8 numbers that answer questions like:
- How much knowledge is actually verified, not just claimed
- How much momentum the building has built up, and which way the trajectory is trending
- How intense the work has been
- How fast new vocabulary is being learned
- How dense and how abstract the domains are
- How persistent the practice is
The second model reads habits. It turns the sensors into 10 numbers about rhythm and retention:
- Whether the work comes in bursts or stays consistent
- Whether the domains are specialised or balanced
- How stable the output intensity has been
- How much mastery is shared across skills
- How strong the growth trend is
- Whether skills get retained and revisited, or dropped after the first use
Together these 18 numbers are the career DNA. They describe not just what is known, but how the knowledge was built and kept.
Layer 3: Rank Inference
The career DNA is then compared against eight archetypes, one for each rank from Academic through Principal. Each archetype is an 18-dimensional vector that describes what a career at that stage typically looks like in both shape and structure.
The comparison uses cosine similarity, which measures the angle between two vectors. A small angle means the career shape closely matches that archetype. A large angle means it does not.
But similarity to an archetype alone is not enough. A year-old career that happens to look structurally similar to a senior engineer should not be called senior. So the similarity score is multiplied by a Gaussian function centered on a tenure anchor for each rank. The Gaussian peaks when the career length matches what that rank typically requires, and decays naturally as the distance grows. There are no hard cutoffs. The math does the gating.
The rank with the highest combined score wins.
What the Breakdown Percentages Mean
Every rank shows a percentage, not just the current one. The percentage for the current rank reflects how strongly the career DNA matches that archetype given the current tenure. The percentages for lower ranks show how completely those stages have been transcended. The percentages for higher ranks show how far along the trajectory is pointing toward them.
A score of 76 percent for Professional and 44 percent for Senior does not mean you are 44 percent senior. It means the structural shape of the career is already 44 percent aligned with what a senior career looks like, and the tenure Gaussian will close the rest of the gap as the career grows.
The 12 Fields
The same three-layer pipeline runs separately for each of the 12 engineering disciplines:
- AI and Intelligence
- Systems Architecture
- Infrastructure and Operations
- Data Engineering
- Security
- Core Logic and Algorithms
- Product and Strategy
- Craft and Design
- Web and Frontend
- Mobile
- Quality and Reliability
- Platform and Developer Tooling
Each field gets its own sensors computed from only the projects that belong to it. That means a strong AI background does not inflate the Infrastructure score. Each field earns its rank independently.
What the Rank Actually Means
The goal is not to game a metric. It is to have a metric worth taking seriously.
If the rank says Professional, it means the shape of the actual work, the retention of skills, the density of domains, the growth trajectory, and the execution consistency all point to that stage. Not because someone spent a certain number of years at a company, but because the evidence adds up that way.
That is the only version of a career level that means anything.