Real-Time AI Tutoring: How Drillby Keeps Students Engaged Online
Student engagement is vital to the tutoring process.
Professional tutors know this.
However, a problem arises when a lot of this communication goes unspoken. The tutor needs to read the student’s reactions and sense the student’s struggles as the lesson goes on, particularly in order to get more out of the student when there’s evidence of confusion, lack of engagement, or frustration.
Any tutor who was tutoring before the Zoom boom understands that reading student reactions is much easier in person than it is online. Having the student on camera is one thing, but the stress of juggling a whiteboard, documents, a lecture, and keeping an eye on the window with the student’s face lends itself to fatigue.
That’s why teachers are susceptible to Zoom Fatigue perhaps even more than students.
Enter Drillby’s Real-Time Analysis
Real-time AI class analysis is one of the major features that separates Drillby from other online tutoring platforms.
Drillby’s real-time class analysis gives extra backup, helping the tutor focus on the lesson by providing the tutor a constant reading of the student’s engagement level.
- Discussion Content Tracking
Drillby’s AI tracks the content of the classroom discussion in real time.
The AI is aware of the class topic, and keeps an eye on when topical discussion is high, low, or the class has drifted off-topic. Then, the AI provides the tutor a notification at the top of the screen (invisible to the student), so the tutor has that extra nudge to get the conversation back on-topic.
Given that establishing rapport with students is of primary importance, it’s understood that class time won’t be 100% devoted to the topic at hand. Percentage tracking of on-topic and off-topic conversation also helps the tutor to understand optimal amounts of tutoring time versus unrelated discussion.
- Contribution Tracking: Student and Tutor
In cases where the student is not contributing well, the AI will provide a banner to the tutor noting the lack of engagement, contribution, or a warning about the student’s perceived emotional state. This is particularly useful because it’s fairly straightforward, on video calls, to track a speaker’s emotion.
What other platforms lack, however, is a mechanism to track the non-speaking members of the call. This is a simple fix that provides great benefit to the tutor as she conducts a lesson.
In cases where the tutor is addressing other PDFs, working on the whiteboard, or speaking with other students, this provides a useful notice that a student might be losing the thread of the conversation.
- Live Document Selection
Discussion within the classroom can also help the tutor by giving the AI an idea of relevant documents to cue up.
Based on the content of the lesson utterances, the classroom will be able to pre-select relevant documents for sharing. These will be noted at the top of the tutor’s selectable files.
This saves valuable lesson time that tutors spend hunting through their entire file libraries, which–even when well-organized–can number in the hundreds or thousands of documents. This saves having to hunt around, pawing through folders and search bars, all the while trying to keep students engaged.
In essence, the AI acts as a teaching assistant who can bring the correct document up in real time and even suggest further reading when appropriate.
- After-Class Analysis
The classroom transcript is able to associate each utterance with emotions, so the tutor can review and understand where the student became confused or frustrated in more granular detail than might be available in the moment.
This will also allow tutors to see which students are less active in class, and which topics improve engagement among students overall. This allows for better analysis of what’s working and makes suggestions about how the tutor can keep students’ attention at as high a level as reasonably possible.
Furthermore, the suggested content and scheduling of the next lesson will be registered based on discussion during the current lesson. Further reading, homework, and more can be recommended for the student with the tutor simply needing to adjust, then sign off before sending to the student.
- Classroom Replay
The classroom replay allows the tutor to observe similar metrics of the students’ attention laid over the video of the session.
In this way, the tutor can see a live replay of the specific activity in the classroom when the students were most and least engaged and can adjust future lessons accordingly.
- Video and Document Storage
In case you’re worried about how much space this all takes up, Drillby will scale up the basic storage.
At the time of writing, the $29 monthly plan, first to be released, is aiming for 50gb of cloud storage per tutor. The different tiers will be released over time: the $59 tier is currently slated to provide 80gb of storage, while the $119 plan will provide 100gb of storage.
This includes all of the session videos, which will be available for up to 60 days after a given lesson, storage of any evergreen video content, and the libraries of documents and videos that a tutor needs for daily teaching.
Classroom Metrics – Data to Improve, Not Punish
Tutoring is a job that involves soft skills as well as hard skills. There are just certain things that it’s hard to turn into metrics.
That doesn’t stop a lot of larger tutoring platforms, and even some funded agencies, from trying to quantify everything. In cases such as these, tutoring becomes more about getting five-star reviews from students (sometimes lesson-by-lesson), increasing student engagement, or even curious ideas such as rating “student confidence.”
Most tutors have had the experience of preparing a student well–very well, in fact–for an exam only to have the student underperform due to illness or crippling anxiety in the testing room. For the student, it’s devastating. For the tutor, this is existential: a data point that an agency can use as a cudgel.
As we digitize our world further and further, it becomes clear that the more metrics are available, the more that people work toward improving metrics–often a distraction from the changes that could make actual improvements for students.
Metrics the Drillby Way
Vitally, Drillby keeps this information specific to the classroom. A tutor can use these metrics to adjust lesson delivery in ways that improve student engagement as well as a tutor’s own experience.
For example, a tutor might intuit that student engagement is lower on Fridays. Understanding this, the tutor can propose less lecture and more interaction to improve engagement in Friday afternoon lessons.
However, there are a couple of concerns worth identifying. First, this is meant to give information, not prescription to the tutor. Drillby isn’t TikTok, and the information is not to encourage tutors to focus on the trendy or on virality.
Second, all such metrics remain private and within a single classroom. Tutors who have been burned by other tutoring platforms might be concerned that this system can be used to determine engagement metrics or set KPIs that the individual tutor would be responsible for meeting.
Importantly, Drillby’s platform keeps the information solely with the concerned parties: the tutor’s metrics stay with the tutor, and the student’s metrics stay with the student. Rest assured, these metrics are always used for the tutor’s self-improvement rather than a larger company’s bottom line.
Even more, the information is provided as guidance, not prescription. Having more useful information can be helpful, and tutors are free to do as they please with such information. If it helps you, as a tutor, do your job better, great. If it seems incidental, you’re free to leave it.
Conclusion
It’s a tutor’s job to see that the student’s needs are met.
Many of those needs come in forms that have to do with increasing communication and simple rapport with students. These are the sort of things that even a couple of years ago it simply seemed out of bounds for a computer to be able to recognize.
Modern AI systems, however, are becoming extremely adept at reading emotions. Drillby’s AI provides suggestions of how its perceived emotions can help guide the rapport between tutor and student.
Drillby understands that tutors, as well, have their own needs. The Drillby way is to discuss with tutors and create a platform that provides tutors with the tools and opportunities that real, working tutors need to run a successful online practice.
A tutor-first platform, Drillby is committed to providing a best-in-class platform, software, marketplace, and metrics.
Drillby’s release is slated for 2027, but you don’t have to wait that long. Drillby Early Access is a one-time investment of $29 that gets you access to Drillby’s features as they go live.
Drillby provides tutors a way to make a solid living–without the platform tax.
Are you a tutor who’s tried different platforms and is looking for something better? We want to hear from you!
Be part of the development. Help shape the future of Drillby by discussing your needs as a tutor. Schedule a discovery call here.
Want to learn more about Early Access? Sign up here:
