AI for Learning: OpenClassrooms
How OpenClassrooms’ is adopting AI to help develop, deliver and refresh its vast course catalogue.
👋 Matt here! This week is another guest post from Marion, who spoke to Laura Besnier, Senior Learning Designer at OpenClassrooms. Laura is currently leading the implementation of AI in the Learning Design Team. She shares how AI is influencing her tasks and role more generally and gives her recommendations to get started with AI.
OpenClassrooms is a pioneering online education platform committed to making quality education accessible to everyone, everywhere.
With a mission to empower students through flexible and comprehensive learning paths, OpenClassrooms offers a wide range of courses, from technical to soft skills, designed to meet the evolving demands of the global job market. Their catalogue contains over 600 courses in both French and English. This creates a massive need for regular maintenance due to the majority of the courses being on technical topics.
Transforming course development
The integration of AI into the learning design process at OpenClassrooms has been transformative, reshaping the way courses are developed, structured, and delivered.
One of the early changes brought by AI is the automation of routine content creation tasks.
"For all the recurring content like chapter summaries and course introductions, we now use ChatGPT,” Laura explains. “These were the first tests we did, and we found it to be very effective."
This shift has allowed the team to focus more on the high-level course design and creativity, rather than getting bogged down by repetitive tasks.
AI has also played a crucial role in scenario creation and project design.
Laura says, "I use it a lot for scenario creation in projects, as Chat GPT is great at imagining how to find a common thread and design the entire course around it. For instance, in the last AI course I developed, I heavily relied on Chat GPT to help conceive all the artefacts for the activities and exercises."
This approach has not only saved countless hours but has also enhanced the quality and consistency of the content.
Testing, refining: from quizzes to storyboarding
AI tools have significantly improved the efficiency of various stages in the learning design process.
"For quizzes, we use AI as well,” says Laura. “It’s not yet perfect for highly practical questions, but it's great for knowledge-based queries."
This allows the team to generate a substantial amount of quiz content quickly, which can then be refined and perfected. “Generally, it gives us a solid foundation to iterate on, and allows our focus and our experts focus to be on the most challenging tasks”.
The team has experimented with AI for storyboarding, though this remains a challenging area. "We tested AI for storyboarding, but it’s not great at conceptualising visuals for us yet," Laura notes.
Despite these limitations, AI still aids in generating initial ideas that the team can build upon.
Collaborating with experts
Working with Subject Matter Experts (SMEs) remains a critical component of course development at OpenClassrooms.
To build course scaffolding, they work with both the main SME and a technical reviewer, to get a second pair of eyes.
Laura explains how AI has facilitated this collaboration: "We tried using ChatGPT with our SMEs to streamline the scaffolding process, where experts provide the detailed framework of the course. At scaffolding creation level, it’s not good enough for now: the level of granularity that we look for is not yet met by AI. So we started using it as a technical reviewer instead, who can observe the inconsistencies, or point out what is missing or could be improved.”
This iterative process ensures that the content is both accurate and engaging, leveraging the expertise of human professionals while benefiting from AI’s efficiency.
Streamlining content localisation and accessibility
Localisation of content has also benefited greatly from AI integration.
"For all the visual content, we can give the visuals to Chat GPT, which extracts the text and translates it for us,” Laura explains. “This process takes just seconds, and we already have the translated content. We also tested multi-language AI video tools to sync translation, lips and voices in no time”.
This rapid translation capability allows OpenClassrooms to offer more courses in both French and English much more efficiently than previously. In the future, they plan to use a similar process to expand to offer more languages.
Another use is to providing alternative descriptions for images. This helps meet accessibility requirement and support screen readers.
Recommendations for getting started with AI
For those new to AI, Laura offers some practical advice: "Have ChatGPT or similar tools open and start by asking questions you face regularly in your work. It helps you understand what AI can do for you and how to communicate effectively with it."
She emphasises the importance of continuous experimentation, adding, "Test everything. The more you experiment, the better you’ll understand its capabilities and how to use it to your advantage. At first I was frustrated that some requests were not providing great results, but when refining the prompts I could definitely see how I was also getting better at explaining my thoughts”.
Challenges and future directions
Despite the numerous advantages, there are still challenges to address. AI’s ability to understand and apply specific guidelines can be inconsistent. "If you give too many instructions, it seems to get lost," Laura explains. The team continually works on refining their prompts and processes to improve AI’s performance.
Looking ahead, Laura envisions a more personalised learning experience driven by AI.
"We are considering how to create custom content for learners based on their specific needs and profiles," she says. This would involve using AI to generate course variations tailored to individual learners, potentially revolutionising the way education is delivered. Not tomorrow, but it’s a dream that could technically be realised in the not too distant future for OpenClassrooms learners.
Summary
Here’s Laura’s top tips:
Start to use ChatGPT to handle routine content creation to free up time for more creative work.
Leverage AI to help conceptualise and develop course scenarios and projects efficiently - like a creative colleague that can support and challenge your ideas.
Balance SME expertise and AI results to focus your experts' energy where it is the most critical for your learners.
Employ AI for rapid translation of course materials to offer multi-language content efficiently.
Continuously test AI capabilities in your workflow to understand its full potential and improve your processes. Be resilient and easy on yourself to begin with!
This case study features in Marion’s course Designing Learning in the AI Era. The next cohort starts 21 October.
 | A guest post by
|