Why we built SkillBricks: the credential problem in hiring

For every technical role on the internet, there is one pre-interview signal that matters: the resume. A recruiter scans it in six to eight seconds, an applicant tracking system keyword-matches it in milliseconds, and a hiring manager skims it while the kettle boils. Everything that happens afterwards, the phone screen, the take-home, the panel loop, is triggered by that one document getting through the first cut. If the resume does not land, nothing else does.

This is the credential problem. The strongest pre-interview signal we have in engineering hiring is a one-page summary of employers and education. It over-rewards prestige. It punishes non-linear careers. It is trivially faked, and at the same time trivially misread. For a recruiter with fifty profiles in the queue and no time to verify any of them, the rational move is to filter by surface features: brand-name companies, brand-name schools, years in role, keyword density. That filter leaks qualified candidates every single day, and there is no feedback loop that tells anyone it happened.

The shift

Three things have changed in the last few years that make a different model possible.

First, the pool of qualified engineers without prestigious credentials has exploded. Bootcamp graduates, self-taught engineers, career switchers, engineers from regions where the local stack of top-tier universities just does not exist. These people can do the work. They can stand up a cluster, debug a production incident, write infrastructure-as-code, read a stack trace. They cannot reliably get past the first filter.

Second, it is now technically feasible to put a candidate into a real production-shaped environment in a browser tab. Containers, Kubernetes, short-lived cloud sandboxes, process-level observability. A candidate's shell history and problem-solving path are machine-readable for the first time. You do not have to guess whether someone can work a cluster; you can watch them do it.

Third, recruiters are exhausted. The existing vendor tools (multiple-choice quizzes, algorithm puzzles, long take-homes) either do not predict on-the-job performance or consume so much of the candidate's time that only the desperate complete them. The budget exists for something better. Nobody has built it yet.

SkillBricks exists because those three shifts line up. The thesis is simple: replace the resume's authority with verified evidence from real environments, keep that evidence under the candidate's control, and let hiring teams discover the candidates whose evidence matches their job. Four pillars carry that thesis.

Pillar one: the Wall

Your Wall is your verified identity on SkillBricks, built brick by brick. Each brick is a skill you have proven through real assessment on the platform. Bricks carry a tier (Raw, Solid, Forged, Polished, Diamond) that reflects demonstrated depth, not self-assessment. The Wall is the first thing a recruiter sees, the first thing a peer sees, the first thing you see when you log in. It is your pre-interview signal.

Walls are anonymous by default. Your real name is not on the public Wall, not in the URL, not in the open graph image, not in search results. A system-generated handle labels the Wall until there is an explicit reason to reveal your identity. The Wall is built to speak on your behalf before your name gets in the way of being read.

Pillar two: real-environment assessment

The hard part of assessment is not writing the questions. It is building the environment that lets you ask them honestly. A real DevOps or SRE test needs real infrastructure underneath: a live cluster, a broken deployment, a misconfigured network policy, logs that say what they would say in production. Not a pretend terminal. Not a step-by-step tutorial with a green tick at the end.

SkillBricks assessments run in dedicated Kubernetes namespaces on an isolated execution cluster, one per session. You get a shell, a problem, and a clock. We capture the whole session: every command, every file edit, every moment of hesitation, every pivot. A real-time AI agent observes the session and engages where it makes sense, asking clarifying questions, probing edge cases, and scoring multiple dimensions independently. The output is not a pass-fail number. It is process intelligence: how you approached the problem, what you tried first, what you did when it did not work. That is the signal a hiring manager actually wants.

Pillar three: the candidate stays in control

Candidates are free, forever. No pay-to-take-the-test, no premium tier for better visibility, no algorithmic penalty for skipping the upsell. The platform is funded by hiring teams.

This matters for two reasons. First, it means a candidate's incentive is clean: prove the skill, get discovered, respond or do not respond on your terms. No commercial pressure points at you. Second, it means hiring teams have skin in the game. A team that reaches out has already decided the match is worth pursuing. Contact stays in-platform, which means the candidate keeps a record of who reached out, what they said, and how fast they responded. The old dynamic of cold-sourced LinkedIn spam does not survive that structure.

Pillar four: JD-driven roadmap with live ranking

A Wall is useful. A Wall matched to a specific job is useful and actionable. Every job description posted on SkillBricks is parsed against the skills catalog. Every candidate whose Wall covers that JD is ranked against it, surfaced to the recruiter, and, on their own side of the platform, shown where the gaps are. A missing brick is not a closed door; it is a labeled next step. Complete the missing skill, the ranking updates, and the recruiter sees the new state.

This is the difference between a credential and a roadmap. A CKA tells you what you already did. A JD-matched Wall tells you what to do next, and moves you up the list when you do it. It turns the hiring funnel into a training loop.

Who this is for today

The first version of SkillBricks is built for DevOps, SRE, and support engineers, wedge-first. We chose that market deliberately. It is the population where the gap between CV-based screening and actual skill is largest, where the work happens in real environments we can replicate, and where the self-taught pipeline is strongest. If you are a bootcamp graduate, a career switcher, a support engineer moving into SRE, or an infrastructure engineer from a company nobody has heard of, the platform is built for you first.

Other domains will follow. The assessment platform is designed with pluggable environment adapters so that a Kubernetes live-task today becomes an AWS scenario, a code runtime, or a Linux container tomorrow without rewriting the engine. But the first year of the platform belongs to the wedge market that needs it most, and our content, challenges, and community will reflect that.

Close

If you have ever had a qualified engineer bounce off your hiring funnel because the CV did not read right, or if you are that engineer and you are tired of the funnel bouncing you, SkillBricks is built for both sides of that problem. Build a Wall. Prove a skill. Let the evidence do the first round of talking.

Sign up, start a calibration, and lay the first brick: Create your Wall.