Learn how we increased JFrog's audience for their SwampUP conference videos without the need for professional translators or video production expertise.
Client
JFrog is a leader in DevOps and software distribution, offering a suite of products designed to streamline the continuous delivery and management of software releases. Known for its Artifactory and Xray solutions, JFrog's mission is to enable continuous software updates across industries, promoting faster, secure, and more reliable software delivery. One of their flagship events is the SwampUP conference, an annual gathering that brings together developers, DevOps professionals, and software innovators to explore industry advancements and share knowledge. SwampUP is a critical platform for JFrog to announce new products and partnerships, highlight their thought leadership, and connect with their customers worldwide.
However, JFrog has struggled to effectively engage their international audience during the conference. All conference talks are presented in English, yet a large share of JFrog users are French and German speakers. Engaging these users with English videos hasn't worked well, as many of their French and German customers prefer consuming content in their native languages. Additionally, deep technical terminology in the presentations make it particularly difficult for non-native English speakers to follow.
JFrog sought a way to improve global audience engagement with their French and German customers by localizing the recorded videos from the SwampUP conference. They wanted a more engaging solution than traditional subtitling or audio-only dubbing solutions, one that would retain the energy of their speakers and ensure that the key messages and technical nuances in the content resonated deeply with viewers worldwide. Key to that goal was not only creating translated and dubbed audio, but ensuring the audio sounded like the original speakers and that the lips and body movements matched the audio. In essence, JFrog wanted localized videos that looked and sounded like they were recorded natively in French and German.
Challenges
JFrog presented several challenges in localizing the high-quality, expert-driven content of SwampUP. First, the technical demands of SwampUP content were significant. Conference talks weren't simply single speakers standing and speaking directly into the camera. They included varying face angles, multiple speakers—sometimes interviewing one another on stage— and on-stage movement. Maintaining the natural look and feel of the localized videos in these complex scenarios was crucial to retaining the integrity and authenticity of these talks.
Additionally, JFrog didn't have available video production resources to dedicate to localization. Nor did their team have experience with the specialized video editing software or dubbing tools typically required for video localization. They needed a non-technical, hands-off solution for translating and editing the videos that anyone could use, one that could simplify the process without requiring their team to develop new technical skills or invest in expensive video production tools.
Finally, although JFrog had local business and marketing teams, they lacked a formal localization team to verify and conduct quality control (QC) on translations. Ensuring that the adapted presentations retained all the technical nuances of the original talks was critical to the outcomes they desired. Therefore, they needed a solution that would simplify QC and translation adjustments for non-localization experts.
Solution
Panjaya's AI-powered video adaptation platform, BodyTalk, was the ideal solution for JFrog's localization needs. BodyTalk translates videos, creates voice-cloned audio dubs of the translations, and matches the speakers' lips to the new audio, automatically, without the need for video production expertise or familiarity with complex video editing or dubbing software. Additionally, BodyTalk handles lip-matching and audio/body movement syncing through a range of complex environments and scenarios. It's specifically designed to account for on-stage movement, varying face angles and sizes, and background noise. All of this results in an incredibly natural-looking result that retains the authenticity and impact of the original videos.
JFrog chose Panjaya not only because of our platform's ability to retain the emotional and educational integrity of the original videos, but also because it allowed simple QC and adjustment of the translations by their local business and marketing teams. Our solution provides a suite of simple to use editorial tools that enable non-experts to quickly review and modify translations. This instilled confidence that all of the nuances of the messaging in the original video would be maintained in the adapted videos.
Implementation
The implementation process involved a simple and seamless collaboration between JFrog and Panjaya to adapt six key talks from SwampUP into French and German. JFrog simply uploaded their content to the Panjaya platform, which generated the initial localized output automatically. From there, the local JFrog business and marketing teams were given access to review and adjust the translations as needed, ensuring the technical accuracy and clarity of the content.
Finally, the refined results were sent back to JFrog. The final videos included single- and multi-speaker, voice-cloned audio, meticulously synchronized with speakers’ lips and gestures, ensuring a seamless, natural viewing experience.
Conclusion
Panjaya's AI-powered video adaptation platform is transforming the way global brands like JFrog communicate with their audiences. By providing localized content that maintains the authenticity of the original speakers, Panjaya is enhancing JFrog's global reach, ensuring that key technical knowledge and insights from SwampUP resonate with audiences worldwide. As we continue to refine our platform, we are dedicated to helping companies like JFrog expand their influence, grow their international communities, and deliver content that educates and inspires—without compromise.