Hire engineers who are force-multipliers with AI
— not people who rubber-stamp it
SkillFoundry runs real-world, repo-based assessments and scores how candidates orchestrate AI, benchmarked against your own team — with a session evidence trail behind every score, so you can defend every decision.
- Evidence reviewers can check
- Missing evidence is not counted against anyone
- AI use is not penalized
- Tasks are repo-based
Evidence-based assessment is rolling out to pilot customers.
Why teams choose SkillFoundry
Four things no algorithm-puzzle screen can do
Traditional coding tests grade output on toy problems. SkillFoundry measures how modern engineers actually work — and proves it.
AI-Orchestration Scoring
Copilot Command measures whether candidates command an AI pair engineer or rubber-stamp its output — with a reviewable replay.
Learn moreTeam Benchmarking
Score every candidate as a percentile against your own engineers, so a number maps directly to your hiring bar.
Learn moreWorkforce Intelligence
Point the same instrument at your existing team via the SkillFoundry macOS and Windows desktop agents — consent-gated capture that maps AI fluency and skill gaps over time.
Learn moreCandidate Authenticity
Tamper-evident, behavioral proof of who did the work — to defeat proxy candidates and identity fraud.
Learn moreEvidence-based assessment
See how the work was done, backed by evidence
SkillFoundry looks past whether the tests pass to how a candidate investigated, used AI and verified their work. Reviewers can check every result against the evidence, and missing evidence is never counted against a candidate.
How they investigate
Whether the candidate took the time to understand the problem and the code before changing it.
How they use AI
AI use is not penalized. What matters is the judgment a candidate applies to what the AI produces.
How they verify
Whether the candidate checked that their solution works, alongside the result of grading the code they submitted.
Rolling out to pilot customers, with a readiness review before any organization's results change.
From Invitation to Decision, Without the Busywork
Your team defines the bar once. The platform handles everything between sending the invite and reading the report.
Assign a real-world task
Pick from your task library or build your own: scoped, ticket-style work in a real codebase, with tests and acceptance criteria attached.
Candidates work, evidence accrues
Candidates complete the task in a real development environment, and evidence of how they work is captured automatically.
Review a defensible report
Each submission is evaluated automatically. You receive a score report with the full breakdown — and your engineers only meet the candidates worth meeting.
Why Hiring Teams Switch
Traditional technical screens
- ✗Senior engineers spend hours writing, running, and grading take-homes
- ✗Algorithm puzzles that don't resemble the job — and candidates know it
- ✗A single number with no evidence behind it when a candidate disputes a decision
- ✗Inconsistent, reviewer-dependent grading across candidates
- ✗No visibility into how the work was actually produced
Assessments on SkillFoundry
- ✓Assessments run, evaluate, and report themselves — no engineer time until the shortlist
- ✓Real-world, repo-based tasks scoped like the tickets your team ships
- ✓Every result backed by trusted test runs on the submitted code and a session evidence trail
- ✓One standardized rubric applied identically to every candidate
- ✓Evidence of how candidates investigate, use AI and verify their work
What's Behind Every Score
A number you can't explain is a number you can't defend. SkillFoundry results are assembled from measurable evidence, and reviewers can follow each one back to it.
Trusted Test Execution
Every submission is graded by running the task's tests on the code the candidate actually submitted, in a trusted environment.
Code Quality Review
The submitted code is reviewed for the qualities your team cares about in code review, not just whether the tests pass.
Behavioral Evidence Capture
The session captures evidence of how candidates work, not just what they submit. Missing evidence is never counted against a candidate.
Responsible AI Use, Not AI Detection
Using AI is not penalized. Reports show the judgment a candidate applied to AI output, which is what matters on the job.
Human Review, On the Record
Reviewers can verify results against the evidence and adjust them, and every adjustment is logged with the reviewer, reason, and timestamp, so the record stays auditable.
Consistent Rubrics
The same scoring model evaluates every candidate on a task. No mood-of-the-reviewer variance, no undocumented criteria.
Built for Teams That Answer to Legal, Too
Automated hiring tools face growing scrutiny. SkillFoundry is designed so your assessment process holds up to questions — from candidates, auditors, or regulators.
Session evidence trail
Candidate sessions are recorded as structured events, so any score can be traced back to what happened.
Human oversight built in
Automated scores are recommendations. Structured manual review — with logged adjustments — keeps a human in the loop.
Data protection
Behavioral events, never keystroke content. Evidence is kept 24 months and artifacts 12 by default, assessment evidence is deleted from live systems within 48 hours of a request, and research use requires opt-in consent.
Fair, job-related criteria
Scoring uses only work products and job-related signals — never demographic attributes — supporting your compliance obligations for automated hiring tools.
Give Your Engineers Their Time Back
Run assessments that grade themselves, and make hiring decisions you can explain — to your leadership, your candidates, and your auditors.