How Preply combines AI and human tutors to personalize learning

YouTube AI Channels Products

Summary

Preply, by using OpenAI's AI as a co-pilot, developed the "Lesson Insights" feature, helping tutors automate lesson preparation and personalize teaching, significantly improving efficiency and long-term learner retention.

No content available
Original Article
View Cached Full Text

Cached at: 06/12/26, 01:35 PM

# How Preply Combines AI and Human Tutors to Personalize Learning TL;DR: Preply uses AI as a co-pilot, combining it with the motivational power of human tutors. Through its "Lesson Insights" feature, it automates repetitive tasks, saving tutors more than half their lesson preparation time and significantly boosting long-term learner retention. ## Not Replacing Humans, but Empowering Them As stated in the video: "It's not about replacing humans, but giving them new abilities, sometimes even redefining the nature of work. For Preply, this is a tremendous opportunity to combine what we humans do best — motivation, interaction, and energy in the classroom — with AI components that act as a co-pilot." Preply is the world's largest language learning marketplace, with 100,000 tutors teaching over 90 languages across 180 countries. Humans have an incredible ability to motivate other humans, and teachers excel at this: "They inspire students to attend class, to complete their homework, and essentially maintain the energy in the classroom that is hard to replicate." ## The Pain Point of Repetitive Tasks At the same time, there are many repetitive tasks that tutors don't want to do and learners don't value highly. Examples include creating personalized plans, typing out detailed plans, taking notes, and preparing lessons. These tasks consume a lot of tutor time without directly improving learning outcomes. ## Lesson Insights: From Generic to Highly Targeted Preply developed the "Lesson Insights" feature. "It's essentially a personalized summary that you receive at the end of the lesson. It provides a recap of what was discussed, feedback on what went well and what can be improved, and suggestions on how to keep improving. It's shown to both the learner and the tutor." This feature shifts teaching "from a generic practice to something highly targeted, truly helping learners achieve what matters most to them at that specific point in time." ## Why OpenAI Preply chose to partner with OpenAI for several reasons. First, accuracy. "We provide feedback on your language, grammar errors, and how you speak. So the accuracy of language output is critical for us. We compared several different models and evaluated them against our framework. OpenAI performed exceptionally well on this front." ## A Tutor's Real Experience One tutor shared: "Lesson Insights helps me personalize lessons for my students by showing me where they need to improve. Before I started using Preply's AI features, I would spend many, many hours preparing lessons and assignments. Now, that time has been cut by more than half." ## Adoption Rates and Retention Data "For the Lesson Insights feature, we have over 70% of tutors opting in to use it, and 75% of learners opting in. So adoption is very high in terms of people wanting to use these features. We also see high per-lesson engagement." But the most critical signal is long-term retention: "We see that after more than a year, 70% of active learners continue to use Lesson Insights." ## AI Empowers Engineers to Focus More on Customers The video also highlights the role of the Codex model: "With a powerful tool like Codex, engineers can focus more on the customer. They don't have to spend as much mental energy remembering syntax, fixing errors, or correcting typos. They focus on the overall architecture of the system and solving customer problems with code. As a partner, OpenAI is able to teach us how to use their product in the most efficient way. And who better to teach you than the organization that created the world's best large language model?" ## Summary Preply unleashes the potential of human tutors by using AI as a co-pilot, while dramatically boosting efficiency through automation. The widespread adoption and long-term retention data for Lesson Insights prove that this combination is not only viable but effective. **Source:** How Preply combines AI and human tutors to personalize learning – OpenAI YouTube (https://www.youtube.com/watch?v=-Jm-Bf3g9r8)

Similar Articles

@FlowOpsDaily: One-Person AI Money-Making Project: Creating Course Preparation Service Packages for Teachers/Training Institutions – Not to Write Lesson Plans, But to Integrate Materials, Exercises, Standards, and Version Management. Drawing Inspiration from the Claude for Teachers Product Direction.

X AI KOLs Timeline

This article proposes using AI to create course preparation service packages for teachers and training institutions, focusing on integrating materials, exercises, and version management, and drawing inspiration from the Claude for Teachers product direction.

SchoolAI builds an AI platform that empowers teachers

OpenAI Blog

SchoolAI launches an AI platform built on OpenAI models that empowers teachers with real-time student progress signals and provides personalized learning support through conversational assistants (Dot) and AI tutors (Sidekick), reaching 1 million classrooms across 80+ countries in two years.

@jinchenma_ai: I think a very important approach to learning AI right now is to use AI to become your personal coach. Many people now try to use AI to help with work or answer questions. But going further, we can also use AI to observe ourselves in reverse and uncover more value for our own growth. Over the long term...

X AI KOLs Timeline

The author proposes using AI as a personal coach, discovering cognitive blind spots and strengths through regular review of interactions with AI, to achieve continuous growth.

@XAMTO_AI: This project called DeepTutor, positioned as a "lifelong learning" personalized tutoring platform, is developed by the Data Intelligence Lab (HKUDS) at the University of Hong Kong. It has already received over 27,000 stars on GitHub. Its core design concept is to provide an "agent-native" learning workspace...

X AI KOLs Timeline

DeepTutor is an open-source personalized tutoring platform developed by the HKU Data Intelligence Lab, offering persistent learning context, a traceable memory system, flexible deployment, and multi-scenario learning support, aiming to solve the issues of conversation continuity and advice explainability in AI tutoring.