This study investigated the depth and impact of interaction in asynchronous online discussions among trainees within the framework of the Community of Inquiry (COI) model. A mixed-methods approach was employed, combining Social Network Analysis (SNA) with a COI coding scheme to analyse the nature and complexity of trainee interactions at the end of a professional development e-training programme. Additionally, a performance evaluation tool was used to assess trainees’ learning outcomes. The findings revealed a high level of interaction complexity, and the network diagram visually represented interactions between the trainer and trainees, as well as among the teachers themselves; the highest frequency of edges, 150, was observed among 23 participants. Regression analysis showed that the COI components collectively explained 48.5% of the variance in trainee performance. Among the three presences, Cognitive Presence was the only significant predictor of performance (β = 0.681, p < 0.05), highlighting its role in promoting meaningful learning. In contrast, Social Presence and Teaching Presence did not show statistically significant effects (p = 0.857 and p = 0.825), which may be attributed to the nature of asynchronous training and the clarity of the course structure that supported self-regulated learning. These findings are consistent with previous research emphasising the centrality of cognitive engagement in asynchronous environments, while suggesting that Social and Teaching Presence serve more as enabling conditions rather than direct predictors of performance. Overall, the study confirms the value of the COI framework in guiding the design of effective asynchronous online professional development programmes.