Gaps flagged as they appear - not guessed on day one.
Instead of a broad assessment before someone has done any work, gaps are surfaced from real early performance signals: task quality, review feedback, cycle times and honest self-assessment.
Stage view
The part of the ramp curve that this stage owns.
Everything this stage takes off your plate.
Early performance signal aggregation
Pull structured signals from work already happening - not a static test kicked off before day one.
Review feedback ingestion
Turn review comments and pairing notes into structured gap indicators rather than free text nobody re-reads.
Self-assessment prompts
Ask the new hire what feels shaky - honest self-report is a real signal, not a bureaucratic form.
Gap specificity, not vibes
'Framework depth on X' rather than 'ramping up' - specific enough that a training recommendation can match it.
Confidence rating on every flag
Every gap flag comes with a confidence score - low-confidence signals surface as prompts, not verdicts.
Manager review is first-class
Every flag is reviewable by the hiring manager before it becomes a training assignment.
One hire, day 12 to day 24 - a gap surfaces cleanly.
A concrete run through this stage - not a diagram of arrows.
- Step 01
Signals collected
Task cycle times, review comments and self-check responses are gathered as the hire actually starts working.
- Step 02
Pattern emerges
Two review threads and one delayed handoff point at the same underlying gap - the system flags it, not a person guessing.
- Step 03
Manager confirms
The manager reviews the flag with context and either confirms it or redirects the signal with a note.
- Step 04
Handoff to training
The confirmed gap becomes the input to the training recommendation stage - specific enough to match a real module.
What feeds this stage
- Task cycle time and quality metrics
- Structured review feedback on early work
- Self-assessment prompts from the new hire
- Pairing and handoff notes
What this stage puts on the curve
- A ranked list of specific gaps with confidence scores
- A skill-gap event on the ramp-up curve
- Manager-reviewed decisions on which to act on
- Structured input for the training stage
A word on accuracy
Skill-gap identification depends on real early signal, which takes a couple of weeks to develop enough shape to be reliable. Every automated flag stays reviewable by the hiring manager, and low-confidence signals surface as prompts rather than verdicts. Manager judgment is not replaced - it is better supported.