Case Study: How Cisco Networking Academy Leveraged AI Post-Editing To Accelerate Multilingual Content Delivery

In today’s global marketplace, the demand for fast, affordable multilingual content delivery has never been higher. However, for many organizations, traditional translation and localization methods have caused bottlenecks and proved too costly to handle all their content, resulting in a backlog of content awaiting translation. But all that has changed.

Enter Lionbridge, the leader in language solutions, and Cisco Systems, a technology leader with a social mission. With Lionbridge’s help, Cisco Networking Academy (Cisco’s technology education initiative) leveraged the transformative power of AI post-editing to reach new levels of content delivery speed, scale, and cost-effectiveness.

How does Lionbridge’s automated post-editing solution work? What did Cisco Networking Academy accomplish? Can your enterprise benefit from this approach to overcome its translation challenges? Find out in our case study.

What Makes Lionbridge’s Solution Stand Out?

While Machine Translation (MT) is a familiar tool, Lionbridge’s translation solution uses a multilayered approach, with the global content lifecycle orchestrated by its AI-first platform, Lionbridge Aurora AI™.

The solution uses the best Neural Machine Translation (NMT) engines for initial translation and frontier Large Language Models (LLMs) for post-editing the NMT output. By incorporating robust linguistic assets — such as Translation Memories (TMs), glossaries, and terminology management — along with varying levels of human-in-the-loop oversight, you get a workflow that’s both agile and reliable.

Quality evaluation options range from no human post-edit to full human post-edit, with additional offerings in between; the content profile, the customer’s goals, and the budget drive the ultimate decision.

What Are the Limitations of AI Post-Editing?

Although AI post-editing is a breakthrough, it isn’t perfect. Low-resource languages, cultural details, and domain-specific terminology can challenge even the best models. Also, generative AI can produce false or hallucinated information. That’s why human validation and continuous prompting are crucial to overcoming these shortcomings and making sure content connects with all audiences.

What Does Cisco Have to Say About Its Experiences With AI Post-Editing?

Cisco’s GTS program manager offered insights into automated post-editing during our webinar, “Can AI Post-Edit?” Watch the recording to learn more.

Get in touch.

Ready to translate more, faster, and within your budget using the most advanced techniques available? Let's talk.