They completed the training. Six weeks later, they cannot perform the procedure.
This is not a failure of training. It is a failure of retention. Knowledge decays on a predictable curve — and without measurement, the decay is invisible until the moment it matters.
What is knowledge retention, and why does training knowledge decay?
Knowledge retention is the degree to which a workforce can still recall and apply training after time has passed. Knowledge decays on a predictable curve: Ebbinghaus documented the forgetting curve in 1885, and without reinforcement most training is forgotten within 30 days.
AmpUp Insights measures retention continuously with spaced retrieval check-ins, tracks decay for each person and procedure against custom rubrics, and reinforces knowledge at the point of decay, so retention is verified rather than assumed from a completion certificate.
The research is clear. Most training is forgotten within 30 days.
Hermann Ebbinghaus documented the forgetting curve in 1885. More than a century of subsequent research has confirmed and refined his finding: without reinforcement, the majority of learned material is forgotten within weeks.
This is not a flaw in your training program. It is a feature of human memory. The brain prioritizes information that is used repeatedly and allows information that is not reinforced to decay. This is efficient for the brain. It is dangerous for organizations where the forgotten information is a safety procedure, a regulatory requirement, or a clinical protocol.
The L&D industry has known about the forgetting curve for decades. The response has been remarkably inadequate: annual refresher training. Completion certificates measure exposure. They do not measure retention. A certificate issued 11 months ago tells you that a worker was in the room. It tells you nothing about what they still know.
Micro-assessments at scientifically optimized intervals.
Spaced retrieval is the most well-validated technique in the science of learning. When the brain is asked to retrieve information at increasing intervals, the memory of that information strengthens. Each successful retrieval extends the retention period.
Initial Assessment
Establishes baseline retention and begins the spaced retrieval cycle while material is still fresh.
Interval 1
First retrieval strengthens memory traces. If retention is strong, the next interval expands.
Interval 2
Subsequent assessments adapt to demonstrated retention. Gaps trigger targeted reinforcement.
Interval 3
Long-term retention verified. Each successful retrieval extends the retention period further.
Initial Assessment
Establishes baseline retention and begins the spaced retrieval cycle while material is still fresh.
Interval 1
First retrieval strengthens memory traces. If retention is strong, the next interval expands.
Interval 2
Subsequent assessments adapt to demonstrated retention. Gaps trigger targeted reinforcement.
Interval 3
Long-term retention verified. Each successful retrieval extends the retention period further.
These are not desktop assessments. They are voice-based check-ins that can happen during a break, a commute, or a quiet moment in the workday.
You define what “retained knowledge” means. For each role. Each procedure. Each regulation.
Generic retention measurement tests whether a person remembers general information. Custom rubrics test whether a person retains the specific knowledge that matters for their specific role.
Procedure-Level Rubrics
Can the worker perform this specific procedure to this specific standard? Measured against the documented SOP, not a generic checklist.
Regulation-Level Rubrics
Does the worker retain awareness of this specific regulatory requirement and its application to their role? Mapped to your compliance framework.
Equipment-Level Rubrics
Can the operator demonstrate competency with this specific piece of equipment, including the safety procedures specific to its operation?
Site-Level Rubrics
Does the worker retain knowledge of site-specific procedures, hazards, and requirements? Unique to each location, each project, each environment.
See which competencies are fading — before they fail.
A completion percentage. No retention data. No decay alerts. No way to know what anyone actually remembers.
Retention curves. Decay alerts. Intervention triggers.
Decay tracking turns retention from an invisible problem into a visible, manageable, measurable one.
Voice-based check-ins that verify retention without pulling workers off the job.
The delivery mechanism for spaced retrieval is the Conversational Intelligence agent. The agent initiates: “You completed confined space entry training four months ago. Can you walk me through the atmospheric monitoring procedure?”
The worker responds in their own words. The system evaluates the response against the custom rubric — not for exact wording, but for demonstrated understanding. If retention has decayed, the system delivers targeted reinforcement and schedules a follow-up at a shorter interval.
This is not a quiz. It is a conversation. The result is continuous competency assurance based on demonstrated retention, not assumption.
Learn about Conversational Intelligence →Related Capabilities
Conversational Intelligence
The voice-first AI agent that delivers spaced retrieval check-ins and real-time support.
Learn more →Competency Assessment
Adaptive assessments that measure what your workforce can actually do, not what they clicked through.
Learn more →Analytics
Longitudinal competency data, decay alerts, and team-level readiness scores for every role.
Learn more →Frequently asked questions
Hermann Ebbinghaus documented the forgetting curve in 1885, and more than a century of research has confirmed it: without reinforcement, the majority of learned material is forgotten within weeks. It is a feature of human memory, not a flaw in a training program, but it is dangerous when the forgotten material is a safety procedure or regulatory requirement.
A baseline assessment establishes retention while material is fresh, then retrieval check-ins follow at expanding intervals. Each successful retrieval strengthens memory and extends the next interval, while gaps trigger targeted reinforcement. Later assessments adapt to each person's demonstrated retention.
Yes. Custom rubrics measure whether a worker can perform a specific procedure to the documented SOP, retains awareness of specific regulatory requirements mapped to your compliance framework, can demonstrate equipment competency including its safety procedures, and retains site-specific knowledge unique to each location.
Completion certificates measure exposure, not retention. A certificate issued 11 months ago tells you a worker attended training, not whether they can still perform the procedure today, and without measurement the decay is invisible until the moment it matters.
A voice-first AI agent delivers spaced retrieval check-ins and real-time support: short conversations that verify what has been retained, identify what is fading, and reinforce it, with adaptive assessments and decay alerts feeding team-level readiness data.
See what your workforce still knows.
Not what they completed. Not what they passed. What they retained — measured continuously, verified by evidence, and visible in real time.