Hire and train by skills, not seat time

Skills-based learning organizes training around discrete, demonstrable skills rather than seat-time courses. Instead of 'completed a 40-hour Python course', the credential reads 'demonstrated proficiency in: list comprehensions, decorators, async/await, error handling, type hints'. Each skill is mastery-gated. The shift matters because employers increasingly hire by skills (not degrees) — McKinsey 2024 found 70% of employers now consider skills-based credentials in hiring.

CBE is the academic version (Western Governors, Capella, Southern New Hampshire) — generally tied to a degree program with broader competencies. Skills-based learning is the granular workforce version — individual skills (often 50-200 per role) tracked as a graph. Both share the same DNA: mastery > seat time. Drivia supports both — schools use it for CBE-aligned degree programs, employers use it for skills-graph training and verification.

Each course is decomposed into 20-200 discrete skills. Each skill has prerequisites, a mastery quiz, and a confidence score. As learners progress, their personal skill graph fills in. The graph powers (a) personalized path generation — learner skips skills they already have, (b) verifiable certificates that list every skill demonstrated, (c) employer queries — 'show me everyone who has demonstrated AWS Lambda + IAM + KMS at >85% confidence'.

Yes. Every verifiable certificate carries a unique public URL where employers (or anyone) can confirm the skill list, the mastery scores, and the date demonstrated. No PII leaked. Tamper-evident with cryptographic signing. We're also piloting LER (Learning and Employment Records) export to the T3 Network and a Open Badges 3.0 standard for portability across platforms.