Synchronous communication, characterized by its rapid messaging and immediate feedback attributes, is pivotal in bridging the transactional distance in online education. However, learners are demotivated while facing learning challenges while working in isolation, in the comfort of their place, and at their own pace. Learners are given project-based assessments in an online environment whereby a collaborative setup is privileged. This research delves into the intricate dynamics of tutor-learner interactions during online tutoring sessions, aiming to uncover the types of support learners seek in a distance education setting. Employing content analysis, the Activity Theory of Engestrom is employed to categorize and qualify the diverse interactions. The interactions among four tutors and their cohort of around 11 online learners are scrutinized, resulting in the analysis of 3199 lines of interactions. The findings underscore the richness of interactions over time during the chat sessions, wherein the tutors provide immediate responses and personalized guidance as per the needs. Learners demonstrate autonomy in their inquiries, yet uncertainty is also evident. It was found that distinct types of supports were identified with a high percentage related to non-cognitive queries. Therefore, this research recommends an augmented approach and a new strategy, wherein chatbots facilitate higher-level discussions, working synergistically with human tutors to foster meta-cognitive accomplishments among learners. In online education, effective learner support remains a challenge, and this research contributes to the advancement of smarter learning environments that offer enhanced and technologically augmented support to learners with a focus on encouraging metacognitive interactions.