Three engineering organizations that started with codebase signal and ended with measurable upskilling results. Company names are pseudonymous; team sizes, tech stacks, and outcomes are accurate.
~34%
faster time-to-proficiency
3 of 5
incident-linked gaps closed in first quarter
Case studies
Fintech · 50 engineers · GitHub + Jira
Voltcraft Payments reduced new-hire time-to-proficiency by 34% in one onboarding cohort
~34% faster
The L&D manager at Voltcraft Payments had run bi-annual skill surveys for three years. After connecting GitHub and Jira, the competency graph surfaced three gaps in distributed systems error handling — specifically around retry logic patterns and idempotency key management in their payments codebase — that the surveys had never flagged. All three correlated with P1 incidents in the prior two quarters. New engineers onboarded against signal-derived paths instead of the previous generic distributed systems track. The next cohort's time-to-proficiency dropped by 34% measured by time to first unsupervised production deploy.
"The survey told us our senior engineers were strong in Kubernetes. The competency graph showed us they'd never handled the specific failure patterns our infrastructure actually generates. Those are not the same thing."
Cartflow cut P1 median response time by half after observability upskilling paths reduced on-call bus factor
P1 MTTR ↓ 51%
Cartflow's infrastructure team was seeing repeated P1 escalations in their observability layer. PagerDuty records showed that 80% of incident involvement was concentrated in two engineers. GitHub contribution history confirmed that only those two had hands-on exposure to the Prometheus + Grafana stack the team ran. Tunlai generated targeted upskilling paths for the 14 engineers with lowest observability coverage. Within one quarter, P1 median response time dropped 51% as the on-call rotation became genuinely distributed.
"We knew we had an on-call bus factor problem. We didn't know it was a skills gap problem until the incident involvement data showed us who had actually touched the relevant infrastructure under pressure."
B2B SaaS · 22 engineers · GitHub + Confluence
Praxelo identified two cloud migration skill gaps from PR data before a multi-region rollout — closed in 8 weeks
8 weeks to close
Praxelo was migrating their SaaS product to multi-region AWS infrastructure. PR review patterns showed that most engineers lacked hands-on experience with VPC peering, Route 53 failover routing, and Aurora Global Database replication — knowledge gaps that hadn't surfaced in surveys because engineers had read the docs and assumed familiarity. Confluence contribution data confirmed the knowledge existed in documentation but not in the team's hands. Signal-derived paths closed both gaps in eight weeks. The migration completed without an escalation.
"Eight weeks from 'we have a gap' to 'the team handled the migration without escalating to me once.' I'd never been able to say that before a major infrastructure change."
What teams say
Straight from engineering leadership
"I've tried three L&D platforms in the past four years. This is the first one an Engineering VP has asked me to expand instead of cut during budget season."
"Our engineers were skeptical about anything that touched their GitHub. The read-only setup and the fact that it immediately showed them relevant gaps — not generic courses — changed that fast."
"The Skill Gap Analytics dashboard gave me something I could show a CTO in 30 seconds. Not a spreadsheet of completion rates — actual skill coverage vs. codebase demand."
Want results like these?
Connect your first integration and see your team's first competency gap within days.