Aptem Enhance is the AI feature set within Aptem Apprentice, designed to reduce the administrative burden on coaches, give learners faster and more consistent support, and help training providers build a delivery model that is both high quality and financially sustainable. This article explains how the four core Aptem Enhance features, the virtual assistant, checkpoint, marking aid, and enhanced reviews, work together across the full apprenticeship delivery lifecycle, from the initial onboarding stage through to progress reviews.
The problem these features are designed to solve
Most training providers operate under real and sustained financial pressure. Apprenticeship funding rates have not kept pace with rising delivery costs, and providers are being asked to do more, and do it to a higher standard, with fewer resources. The result is that coaches, who represent the majority of headcount in most organisations, are spending a significant portion of their time on administrative tasks: marking work, writing up review notes, answering the same learner questions repeatedly, and manually checking learner understanding.
Research cited in the sector suggests that coaches spend between one and two days per week on marking alone. Add in the time taken to write up progress reviews and coaching catch-up notes, and it becomes clear that a large share of a coach’s working week is occupied by work that does not require their professional expertise. The consequence is that learners receive less direct coaching time, feedback is sometimes less detailed than it should be, and the coaching relationship that sits at the centre of good apprenticeship delivery is squeezed.
Aptem Enhance addresses this by embedding AI into the workflow at the points where it can take the greatest weight off coaches, freeing them to focus on the teaching, mentoring, and contextualised support that only a person can provide.
Using AI features during onboarding and applicant assessment
The first opportunity to use Aptem Enhance features comes before learning even begins, during the onboarding and applicant assessment stage. Aptem Apprentice supports highly configurable onboarding programmes that can be built to walk applicants through eligibility checks, self-assessments, and compliance paperwork before they are formally enrolled.
The skills radar is a self-assessment tool that can be used during onboarding to capture how an applicant rates their own competence across a set of knowledge, skills, and behaviours relevant to the standard they are applying for. If an applicant rates themselves unusually highly in a particular area, that can indicate a need to check whether prior learning adjustments are appropriate, or whether the standard itself is the right fit for them.
Checkpoint, one of the Aptem Enhance features, is available within onboarding programmes as well as delivery. This means that where an applicant has self-assessed at a high level of competence against a specific criterion, an administrator can create a targeted checkpoint assessment for that applicant to complete as part of the onboarding process. The checkpoint presents the applicant with multiple-choice questions mapped to the relevant knowledge, skills, or behaviours, giving the enrolment team an objective indicator of actual understanding alongside the applicant’s self-reported score. A high self-assessment score combined with a low checkpoint result is a meaningful data point for the eligibility review.
Where an eligibility review needs to happen in conversation, Aptem’s Microsoft Teams integration allows a review to be scheduled directly as a Teams meeting from within the platform. If the meeting is recorded, the transcript is pulled back into Aptem automatically. The enhanced reviews feature can then generate a structured summary of the discussion, removing the need for the enrolment officer to spend time writing up notes afterwards. Where recording is not permitted, transcripts can be uploaded manually for summarisation.
Supporting learners through delivery: the virtual assistant and checkpoint
Once a learner is on programme, the virtual assistant is available to them at any time through the Aptem platform. It is contextualised to the learner’s specific programme and apprenticeship standard, which means it can answer questions about the standard, explain knowledge, skills, and behaviours, and help a learner work through something they are stuck on, without the learner needing to wait for their coach to be available.
This is particularly valuable for providers delivering in sectors with shift patterns, such as hospitality, healthcare, or social care, where learners may be doing their learning outside conventional working hours and cannot easily reach their coach with a question. If a learner comes to do their learning at 7pm and gets stuck, the virtual assistant removes the barrier that would otherwise cause them to put the work down and come back to it later, if they come back to it at all.
Checkpoint works alongside the virtual assistant to support continuous knowledge verification. Rather than relying on a single high-stakes assessment at the end of a unit, checkpoint provides a bank of multiple-choice questions mapped to the standard’s knowledge, skills, and behaviours. Questions are generated in advance, quality assured, and organised by apprenticeship standard. The system randomly selects a set of questions for each learner to complete.
Checkpoints can be triggered automatically on completion of a learning activity, on a periodic basis such as monthly, or manually by a coach. When a learner completes a checkpoint and gets a question wrong, the system shows them the correct answer and explains why their answer was incorrect. If they are still uncertain, they can ask the virtual assistant to explain further, with the full context of the question already loaded in. This combination helps learners build genuine understanding progressively, rather than accumulating gaps that only become visible close to end-point assessment.
For coaches, checkpoint results are visible on each learner’s profile and aggregated across a caseload. Before a coaching session, a coach can review which knowledge or skills areas a learner is consistently getting right and which they are struggling with, and use that to shape the session rather than spending time in the session discovering this from scratch. At a provider level, checkpoint data can reveal where all learners on a particular standard are struggling with the same criteria, which may indicate a curriculum gap or a professional development need for one or more coaches.
Reducing marking workload with marking aid
Marking aid is an AI-assisted marking tool that supports coaches in marking written learning plan assignments. It uses classification AI, which means it learns from the way coaches mark work over time, rather than generating feedback from scratch using a large language model.
When a learner submits a written assignment, marking aid analyses the submission against the rubrics and feedback content configured for that exercise by your team. It highlights sections of the learner’s answer, applies positive or constructive feedback from the feedback bank, and suggests a grade based on the rubric. Each suggestion comes with a confidence score. The coach reviews the suggestions, accepts what is correct, overrides what is not, and marks the work as complete. Each time a coach makes an override, marking aid updates its model. Over time, as more submissions are marked, the AI’s suggestions become more accurate and the coach’s active input per submission decreases.
The practical impact is significant. Sector evidence indicates that coaches spend one to two days per week on marking. Even a partial reduction in that time, achieved by removing the need to write standard feedback from scratch on every submission, compounds quickly at scale. For a provider with a large caseload, the hours freed across a team of coaches translate into more time for direct learner support, more capacity for the quality and curriculum work that improves outcomes, and a more sustainable workload for the coaching team.
Saving time on reviews and coaching catch-ups with enhanced reviews
The enhanced reviews feature uses the Teams integration and AI summarisation to reduce the time spent on review write-ups. When a progress review or coaching session is held via Teams, the recording generates a transcript. Aptem pulls that transcript into the review record, and the AI generates a structured summary of the discussion.
The time saving per review is around half an hour. That figure becomes substantial when multiplied across a provider’s full review volume. For a provider with 7,000 learners completing four reviews per year, that amounts to 28,000 reviews annually. At half an hour per write-up saved, that is 14,000 hours of coach time recovered each year. For a provider also using enhanced reviews for monthly coaching catch-up sessions, the saving compounds further. A coach with 45 learners doing monthly check-ins would save around 22 to 23 hours per month on write-ups alone.
The feature is also available for sessions where recording is not possible. Transcripts can be uploaded manually for summarisation, including edited transcripts where sensitive content has been removed before upload.
A delivery model built around the coach, not around replacing them
A common concern when providers first consider AI in delivery is that it will reduce the role of the coach. The evidence from providers using Aptem Enhance points in the opposite direction. Learners who use the virtual assistant and checkpoint arrive at coaching sessions having already explored content, tested their understanding, and identified where they are less confident. That makes the coaching session more productive. The coach spends less time on routine question-and-answer and more time on the deeper discussion, practical application, and contextual guidance that coaching is for.
The design of Aptem Enhance reflects this. Each feature handles a specific administrative or repetitive task so that coaches can focus on what requires their expertise. The AI marks first drafts, the AI summarises meeting notes, the AI answers factual questions at 11pm when the coach is not available. The coach is freed to do the mentoring, the stretch and challenge, and the professional conversation that supports genuine learning and timely achievement.
See how Aptem Enhance works in practice, book a demo or watch the session below.
Frequently asked questions
Are checkpoint questions generated by AI on demand?
No. Checkpoint questions are generated in advance against each apprenticeship standard and go through a quality assurance process before being made available. The system randomly selects from a bank of pre-approved questions for each learner. This ensures the questions and answer options are clear and appropriate, which would be harder to guarantee with questions generated in real time by a large language model. Providers can report questions that appear problematic, and those are reviewed and adjusted centrally.
When can checkpoint assessments be triggered?
Checkpoints can be configured to trigger automatically when a learner completes a specific learning activity, to run on a periodic basis such as monthly or every two months covering all criteria completed up to that point, or to be created manually by a coach for a specific learner. Periodic and automatic triggers are configured in the Aptem settings area by administrators with the relevant permissions. Checkpoint is also now available within onboarding programmes, so it can be used as part of applicant eligibility assessment before a learner formally starts on programme.
What if our organisation does not allow Teams meetings to be recorded?
The enhanced reviews feature can still be used without a recording. If a transcript is generated through another means, or if a transcript file is available, it can be uploaded directly into the review in Aptem and summarised using the AI feature. Where a transcript contains content that should not be included in the summary, that content can be removed from the file before upload.
How does marking aid learn, and does it share data between providers?
Marking aid uses classification AI, which learns from the marking decisions made by coaches within your organisation. It does not use a large language model to generate feedback and it does not share learning across different providers. The AI’s understanding of what constitutes appropriate feedback and grades for a given exercise is built entirely from your team’s own marking, which means it reflects your organisation’s standards and voice.
How do learners respond to AI-driven support tools?
Learners who use the virtual assistant and checkpoint regularly tend to arrive at coaching sessions better prepared, which makes the sessions more valuable for both the learner and the coach. The convenience of having support available at any time, combined with the lower-pressure environment of interacting with a tool rather than a person, appears to help learners build confidence. The AI features complement the coaching relationship; they do not replace it.
