Online course creation through AI: how we increased engagement, accelerated production, and streamlined SMEs involvement

Apr, 25, 2024
Tim Aleksandronets
Tim Aleksandronets
CEO at Blue Carrot


High-priority and time-sensitive e-learning projects call for innovative solutions. With the development of AI tools for course creation, meeting greater demands becomes possible. This article discusses how Blue Carrot has saved significant time and money using AI for online course creation while decreasing SME involvement. These tools can generate engaging, rich e-learning courses that maintain a professional look and feel.

E-learning professionals can find a wealth of benefits in AI tools without reducing the quality of the product (In fact, it can actually enhance it!). We highlight one of our recent projects to demonstrate the high return on investment when creating an online course using AI. 🧐


  1. An Introduction to the Practical Use of AI in E-Learning Development
  2. Practical Case: E-learning content localization in multiple languages with a tight timeframe
  3. Generative AI for Video Creation: Our Solution to Rapid E-Learning Production
  4. How We Did It: Our Strategy for AI Course Creation
  5. Lessons Learned
  6. Generative AI Has a Logical Place in the Future of E-learning
  7. Blue Carrot Is Your Trusted Partner in AI for E-learning Content

An Introduction to the Practical Use of AI in E-Learning Development

With increasing market demand for e-learning services, e-learning professionals must find strategies for streamlining and expediting content development without sacrificing quality. A growing number of AI tools — such as ChatGPT and other NLP-based AI tools, or Generative AI tools like Synthesia, HeyGen, and others — enable rapid creation and localization of e-learning content, reducing the time taken from weeks to just a few days in some cases. 

Learning engagement is significantly increased through AI’s support in producing high-quality captivating media with localization for a global reach. Moreover, using AI minimizes the dependence on Subject Matter Experts (SMEs) by streamlining tasks such as content curation, scripting, assessment creation, and feedback analysis while avoiding complicated and expensive video shooting. Cost savings are realized through the reduction of manual tasks, faster content production, and the ability to update or localize learning programs without significant increases in labor.

Using AI, Blue Carrot has been able to supercharge the production of e-learning content. In the past two months, we engaged with one client to produce more than 70 hours of learning content in four languages. We are pleased to highlight this success story in which we took the opportunity to create an online course using AI with great success and client satisfaction.


Practical Case: E-learning content localization in multiple languages with a tight timeframe


The increasing demand for skilled professionals in the sustainable energy sector drove our customers to extend and localize their online course portfolio to penetrate new markets. The customer required the production of new e-learning content that would reach learners in different countries across four languages. This project aimed to produce 70 hours of e-learning content in 4 languages within 2 months.

🔹 Challenges

  • The tight turnaround time of this project required rapid and resource-effective development and localization of accessible e-learning content to meet global educational needs.
  • It was necessary to produce this content while maintaining top-notch media quality.
  • The courses needed to be efficiently delivered across different linguistic and cultural contexts.
  • The e-learning content had to be “customer-ready” so our client could send it directly to foreign learners without any additional review.

With these requirements, production using traditional production methods was not likely to wrap before the deadline. Thus, we turned to generative AI to achieve the desired results. In this project, we tackled some of the most common modern challenges of online education with the assistance of AI course creation tools.

🔹 Key Objectives

  • Enhance learner engagement through producing high-quality captivating media with localization for a global reach.
  • Reduce dependence on Subject Matter Experts using Generative AI for synthetic videos.
  • Ensure budget and time-to-market efficiency by leveraging generative AI technologies and distributed teams.

Generative AI for Video Creation: Our Solution to Rapid E-Learning Production

We elected to use generative AI for course creation due to its clear benefits and the growing evidence of the effectiveness of synthetic video for e-learning. Generative AI represents a groundbreaking approach to artificial intelligence, where systems are designed to create content that mimics human-like outputs. This technology uses sophisticated algorithms and machine learning models and can generate texts, images, and videos that are often indistinguishable from reality.

📌 Our Solution: Synthetic Videos with AI-Generated Avatars

For the current project, we used generative AI to produce videos that used a synthetic avatar for pre-recorded courses. Using generative AI for video production significantly reduces the time and resources needed to create new content, allowing for a more efficient content creation process.

👉 Take a look at our previous work for an example of high-quality AI-generated video content:

📌 The Evidence for Synthetic Video in E-learning

The efficacy of AI-generated content for e-learning is highlighted in a white paper (Leiker, D. “Generative AI for Learning: Investigating the Potential of Synthetic Learning Videos.” ArXiv.Org,  2023)  from the 2023 International Conference of Artificial Intelligence in Education. The study examines learner perceptions of traditional versus AI-generated synthetic videos.

This paper presents compelling evidence that there are no significant differences in how learners engage with and perceive the educational quality of human-generated synthetic videos in e-learning environments. 👨🏻‍🎓

These findings suggest that AI-generated synthetic learning videos have the potential to be a viable substitute for videos produced using traditional methods in online educational settings. Thus making high-quality educational content more accessible across the globe.

animated men images

percentage agreement content

📌 The approach that we turned down: AI Dubbing

Another emerging technology called AI dubbing is also an option for localization projects that require the same content to be presented in different languages. It can be used as an alternative to the generation of synthetic avatars for pre-recorded courses. The current challenge of this technology is its limitation in making edits to content, such as adding additional phrases to a video. While we did not apply AI dubbing as a solution for this reason, it is a promising method that could further reduce e-learning content development time as AI capabilities evolve.

📌 5 Reasons Why We Chose Generative AI

The benefits of using AI for course building extend far beyond being able to produce high-quality products, although this is certainly a top appeal. Generative AI also provides various opportunities to save time and money while increasing learner motivation and interest with more engaging content. These are some of the primary ways that AI tools for online course creation can maximize the return on investment and client satisfaction of e-learning projects:

  1. Less SME dependence: When using avatars in synthetic video, it is no longer necessary to film SMEs for content production. This is a time and cost-consuming stage that can delay production and limit resources. With the help of AI-powered online course creation, it is possible to create an AI-avatar of the course SME and generate their voice in another language.
  2. Engagement of localized content is better
  • Comprehensive translation of all written content: When localizing content, AI tools allow for changing text on any images used on a video, changing overlay texts on screen, and adding subtitles. These features provide all language learners with an improved experience.animated man and woman with text content in diffent languages
  • Comprehensive voiceover dubbing in multiple languages: Having voiceover in a target language facilitates learning better than only having subtitles.
  • Lower production costs: Voiceover dubbing and AI-generated videos cost significantly less than a live shoot. (Note that while dubbing is cheaper overall, it won’t bring all the other identified benefits of AI-generated videos).

  1. Increased engagement of content: Synthetic videos using avatars are more engaging than any other medium with a comparable production cost including presentations, rich texts, audio, and graphics.
  2. Easy content updates: Generative AI makes it easy to update courses when needed. These changes apply to any aspect of the course including voiceover, text on screen, and content order, and can be implemented even after the courses are shipped to the client. We prepare all the graphical assets in a way that allows instructional designers who may lack motion graphics expertise to autonomously implement changes. Once the edits are finished, re-uploading the video on a hosting platform completes the process. Some platforms like Vimeo allow re-uploading without changing the URL and you also won’t need to change the video link on Articulate slides.
  3. Reduced costs and time-to-market: Using AI for online course creation is significantly cheaper, faster, and easier than doing a live shoot. The resources for one hour of synthetic video production start from 30 hours per 1 e-learning hour and grow from there depending on the complexity of the graphics. With some moderate costs to use the necessary AI apps, synthetic video is slightly more expensive than dubbing, but in some cases comparable. With Generative AI capabilities, it’s possible to produce or localize videos for a course in less than one week.


How We Did It: Our Strategy for AI Course Creation

Here’s how we integrated generative AI into the content development process to enhance production speed and product quality.

✔️ Production Approach Using Synthesia

Blue Carrot chose the AI video production tool Synthesia for the rapid generation of videos and further localization. Synthesia can generate custom avatars, create scenes with multiple avatars to simulate conversations and scenarios, generate natural-sounding voices, clone voices, and produce translations in more than 120 languages. It also has a robust suite of video editing capabilities.

synthesia app interface

Before beginning this project, we re-created live-shot videos in English to AI-generated English videos and used this library as the basis for further localizations.

We also used Articulate Storyline as an authoring tool for the lessons. The resulting interactive slides include the educational text and graphical content, videos, assessments, and links to additional reading materials.

✔️ Localization Approach

We used various efficient techniques to adapt the content for multiple languages and cultural nuances, ensuring seamless delivery of educational courses to the learners.

Our “customer-ready approach” eliminates the need for a client’s review, allowing us to send the translated materials directly to learners upon delivery.” 


Tim Aleksandronets
  • Blending top-quality teams: We combined linguists and developers/editors into one team to streamline the localization process. Our linguistic team consisted of “unicorn” specialists with professional certifications and 5-15 years of proven experience (part of them including practical non-linguistic work experience and engineering education) in the energy domain.
  • Defining localization strategy: The customer took ownership of determining an approach to cultural/local localization. The countries pertaining to the project are all in the European Union and thus use the same currency, metric system, and other commonalities. They decided that translation only needed to be on our end, so we did not have to change references, case studies, or other contexts. We did a comprehensive complete localization of all texts and audio including translating VO, subtitles, texts on videos, texts on images, texts on slides, and others.
  • Applying relevant standards: We used ISO 17100:2015 basic principles as foundation and enhanced some for a more precise content translation, in particular:
    • Having four rounds of internal inter-functional reviews;
    • Reviewing not only text but also output media;
    • Involving subject matter experts for consultancy and systematic quality check.

Example of completed lesson in English

Example of the same lesson in German

✔️ Implementation

With the above plan in place, we implemented these strategies into development, team formation, production, and quality assurance. We used the following steps in our development approach to create an online course with AI:

  • Unified visual style creation for videos and interactive slides, creation of graphical templates;
  • Translation and editing of all texts in CAT;
  • AI Video production in the Synthesia app;
  • Interactive slides production in Articulate Storyline;
  • Comprehensive inter-functional QA after each stage and substage;
  • SCORM packaging and final delivery of the packages and source files.

👉 We ensured that we involved the most qualified and essential team members for the project. Our team included:

  • 1 delivery manager;
  • 2 project managers;
  • 1 translation technical lead;
  • 11 translators;
  • 4 lead editors;
  • 4 reviewers;
  • 3 scriptwriters;
  • 2 graphical designers;
  • 4 Storyline developers;
  • 2 video editors;
  • 3 QA specialists.

We also employed careful measures to ensure the content met high educational and linguistic standards.

  • Technical QA checks: We performed a comprehensive range of technical QA checks (both automated and manual) as part of standard delivery. These QA checks included a review of translated texts and final media files (internal and on-platform).
  • Localization evaluation: We implemented a localization quality evaluation process with native subject matter experts using our DQE matrix. The table below demonstrates the matrix we used to ensure that the linguistic team constantly gives top-quality output.

table with content

✔️ Challenges

Throughout production, as with many projects, some obstacles encountered had to be overcome. Some of our most pressing challenges are detailed below with our approach to resolving them.

  • Consistency in translations: We used multiple translators to reflect the scope and time of the project. To avoid inconsistency between translators, we created glossaries and style guides, conducted team feedback sessions, shared translation memory, and provided additional specific instructions for aligning translations. Translations were all edited by the lead editor and then passed through several automated and manual QA checks.
  • Dropping AI machine translation: Ultimately, using AI machine translation slowed down the process and impacted the product, so we stopped using it.
  • Challenging timeline: With a short turnaround time to complete a sizeable amount of content, we needed to streamline the process between linguists and content developers with weekly agile sprint planning and daily control.

    inno project content sample

    inno project in asan content sample

  • No client review: Our client didn’t have an opportunity to review the foreign content, so our goal was to produce customer-ready content verified for quality.


✔️ Results

We were ultimately able to produce the AI-generated learning content within the client’s designated timeframe and at a desirable level of quality and learning effectiveness. Highlights of our results include:

  • The volume of content produced: With more than 70 hours (1800+ minutes) of learning content, we produced 2500+ slides, 240+ lessons, and 0.8 million words.
  • Language diversity: We achieved the successful localization of content into four different languages: German, French, Spanish, and Romanian.
  • Speed of delivery: Using one of the best AI course creators available, we reduced the production to the client’s designated 2 months. Although we achieved this high-speed delivery, we suggest planning about 3.5 months for a comparable scope to eliminate extreme pressure on the team.


✔️ Impact

Our successful results had a significant bearing on the client’s bottom line and ability to reach their target audience. The courses have been adopted by a broader audience at a steady rate and the initial feedback from learners has been positive. Content is now available for non-English speaking energy professionals in new regions.


Lessons Learned

Here are some of the most important insights we gained regarding the use of generative AI in educational content production and localization:

  • Use of AI:
    • Translation with an AI assistant doesn’t speed up or ease the process when it comes to top-notch output quality.
    • Also, synthetic videos and voiceovers require a linguistic post-production review of media materials.
  • QA process: Comprehensive inter-functional QA with strict control of identified issues and implementation is essential for productions of this scale.
  • Linguistics: For producing top-notch translations, it is necessary to involve subject matter experts as consultants of the linguistic team and reviewers of the final output. It’s also important to discuss expectations for the linguistic part of the work with the client and get approval for the glossary and style guide at a minimum. Additionally, specialized training for translation for Generative AI is needed.

Generative AI Has a Logical Place in the Future of E-learning

Generative AI played a pivotal role in addressing the urgent need for comprehensive and inclusive educational content within the sustainable energy sector. The deployment of generative AI technologies enabled us to expedite content production and ensure high-quality, localized e-learning materials that cater to a wider, multilingual, and multicultural audience. This innovation in e-learning content development and localization exemplifies a significant leap towards more adaptive and accessible learning environments, tailored to the needs of global learners and professionals in the dynamic field of sustainable energy.

Our innovative approach to generative AI, including AI translation and synthetic video production, demonstrates that AI capabilities can revolutionize e-learning content development and localization practices. By integrating AI into our workflows, we overcome substantial obstacles, such as the project’s challenging timeline, reducing dependency on subject matter experts, and the necessity for engaging and accurately localized content.

We foresee more and more applications of generative AI in e-learning productions. The quality of synthetic avatars grows rapidly with visible changes that are noticeably improved every couple of months. We anticipate that these technologies will not only become more prevalent but will also significantly enhance the efficacy, inclusivity, and accessibility of e-learning content across various sectors.” 


Tim Aleksandronets

Your Trusted Partner in AI for E-learning Content

We invite educators, training organizations, and businesses to explore the benefits of AI-enhanced e-learning solutions with us. We’re dedicated to guiding our clients through the evolving landscape of digital learning. Using our extensive media production experience, we provide consultation to help them make informed decisions, such as:

  • Selecting learning design and online learning formats that are most effective for their needs.
  • Implementing the most efficient and suitable formats based on objectives and project limitations.
  • Planning out and executing project development processes.
  • Managing budget and timeline optimization using AI and distributed teams to streamline content creation and localization, without compromising the effectiveness of the learning experience.

Here is the feedback provided for working on the described case study:

Blue Carrot's commitment to excellence content

About Blue Carrot

👉 Blue Carrot specializes in tailor-made e-learning solutions, whether end-to-end online course development with audience research or educational content production and localization. Serving a diverse range of clients for 10 years — from renowned educators and organizations to leading agencies — we facilitate the transition to globally accessible digital learning experiences empowered by AI.


🥕  Check out what Blue Carrot has to offer. 


Reach out directly to our team via the form below and get started on your project immediately! 🔔

Get in touch

Name missed
Email missed
Wrong Email format
Description missed
Rate Article