How a Public Sector Institution Used AI to Spot At-Risk Training Participants

The Challenge

The institution ran training programmes supported by its training operations and in-house learning teams, with trainers and coordinators working directly with participants. Some participants fell behind through absence, low progress or low engagement, and were at risk of leaving a programme before they had completed it.

The difficulty was timing. Signs that a participant was struggling could show up in their attendance, their assignment submissions, how often they used the learning platform or the feedback they gave, but these signs were often noticed late. By the time a trainer or coordinator realised that someone was likely to drop out, the opportunity to step in early had often passed.

The Solution

With cognipeer Studio, our team configured an AI agent that looked across the signals connected to each participant’s progress. It analysed attendance status, assignment submissions, how often participants used the learning platform and the feedback they gave, and identified the participants who showed signs of being at risk of dropping out.

For each participant identified, the agent suggested possible interventions for the trainer or coordinator and prepared personalised reminder and encouragement messages. Trainers and coordinators decided how to act on each suggestion and which messages to use, so the relationship with participants stayed in their hands while the agent pointed them to the people who needed attention.

Implementation

1. Brought the participation signals together. We connected attendance status, assignment submissions, learning platform usage and participant feedback so they could be analysed together.
2. Configured risk identification. The agent analysed these signals to find participants at risk of not completing their programme because of absence, low progress or low engagement.
3. Added intervention suggestions. For each at-risk participant, the agent proposed interventions for the trainer or coordinator to consider.
4. Prepared personalised messages. The agent drafted reminder and encouragement messages tailored to each participant, which trainers and coordinators reviewed before use.

Business Impact

Participants who were drifting away from a programme came to the attention of trainers and coordinators sooner, while there was still time to help. Follow-up became closer, because it started from a clear view of who was at risk, with suggested next steps and messages already prepared. More participants went on to complete their programmes, and the institution made more efficient use of what it invested in training.

• Higher programme completion rates as at-risk participants were identified earlier.
• Closer follow-up of participants by trainers and coordinators.
• More efficient use of training investment, with effort directed to the participants who needed it.
• Interventions decided by trainers and coordinators, supported by suggestions and draft messages.

Technology Behind the Solution