We already
solve wide range
of specific
projects

From hours of original studio recordings of Meky’s singing, we trained a high-quality vocal model. In our creative lab, using AI, we carefully separated the original recording into instrumental and vocal tracks. The cleaned vocal track was reconstructed using the vocal model. The final version of the song features a completely reconstructed vocal track.

A unique workflow for reconstructing preserved DEMO recordings and transforming them into professional studio tracks. Ideal for reviving the works of deceased or long-forgotten artists and their valuable creations.

The main goal of this proof of concept was to find the best balance between output quality and cost. Confirming the scalability of this solution meant staying within 10% of the original production budget while offering a more innovative approach than traditional dubbing. The resulting Romanian version of the commercial features the cloned voice of the Czech actor using generative AI, with approximately 80% lip movement synchronized to the new language track. No CGI was used in the localization process.

 Innovative Workflow for Localizing Costly Ad Productions for Multiple Foreign Markets. Utilizing the Latest Technologies for Emotional Expression of Cloned Voices and Lip Sync Adaptation. Suitable for Localizing a Wide Range of Audio/Video Formats Across Global Languages.

 Fameplay filmed footage to create studio avatars of extremely busy athletes during their intense training periods. To defend their gold medals, they will have no time for filming during the Olympics, making this the perfect situation to use their digital twins, identical in appearance and voice. To ensure that the avatars could interact and behave like regular news anchors and reporters, we filmed material from multiple positions, capturing a variety of gestures, movements, and facial expressions.

One of the first comedic concepts for serious avatar use in television entertainment. The combination of multiple positions and expressions gives viewers the feeling of watching real hosts. Episode production is completely independent of filming, and the ability to scale into multiple languages without dubbing offers absolute creative freedom.

Solution: We mapped out the traditional audiobook production process and broke it down into individual steps. We then designed a new workflow, estimating the time required for each production phase. By carefully selecting tools and creating precise voice models of the authors along with new synthesized voices, we produced the first audiobooks.These were tested on audiobook listeners, and we fine-tuned most of the feedback related to text interpretation and listening comfort.

Our designed and refined workflow is ideal for generating longer texts with emotive narration. A precise approach to voice cloning and synthesis delivers a natural reading performance. Continuous feedback control ensures consistent output quality. The combination of original voice synthesis and emotive delivery is perfectly scalable for localization into global languages.

Fameplay loves creating avatars, improving their technical quality, and discovering new ways to use them. Jiří’s higher-pitched voice with its distinct tone is ideal for cloning. His very first video replaced a demanding trip to Singapore, delivering all necessary messages accurately and clearly in multiple languages.

 Innovative Workflow for Localizing Costly Ad Productions for Multiple Foreign Markets. Utilizing the Latest Technologies for Emotional Expression of Cloned Voices and Lip Sync Adaptation. Suitable for Localizing a Wide Range of Audio/Video Formats Across Global Languages.

The script and creative concept for the video were developed in line with the challenge: What does the year 2030 look like? Can we create a compelling vision of the future in a few minutes without turning on a camera, using only generative AI tools? Voice and likeness clones of top executives were supplemented with visualizations of energy principles for which crucial decisions are being made today. In various scenes, digital twins explained in multiple languages what shared energy, electromobility, and renewable resources will look like, emphasizing ecology and workplace safety even more.

A proof of concept for a video creation approach that completely transforms traditional methods. Utilizing several leading visualization tools revealed current shortcomings but also enormous potential for the future of video production and cinematography. The clarity and impact of the future visualization were enhanced by precise definitions of both applied stylizations.

For an independent producer in the relatively small Czech market, overcoming the language barrier is key to reaching a global audience. At the beginning of 2023, we integrated the first localization tools powered by generative AI capable of understanding the Czech language. To address the various steps involved in localization, we focused on using specific tools (transcription, translation, voice synthesis). Gradually, we refined the workflow for concurrent creation of multilingual translations and involved leading localization procedures integrators and gen AI distribution tools.

Workflow optimized for simultaneous production of multiple language versions. Maximized tool utilization and minimized manual post-production. Proven method to expand audience and increase monetization.

Based on long-term data, we know that making messages more attractive helps convert text to video. Is there a better task for a digital twin than regularly communicating with the public? It never gets tired or sick. So, we created a studio avatar of the director of several departments and a board member, and trained it to convert blog text into a script.

We achieved an efficient and seamless solution for generating and editing scripts based on the company’s blog posts. This approach utilizes and repurposes existing content to increase viewer engagement and intensify the spread of awareness on important, share-worthy topics.

A creative team with over 10 years of experience operating a content platform for the largest Czech search engine, Seznam.cz, is approached with the idea of building a new internet TV for the second largest e-shop in the Czech Republic. The concept is simple: let’s create a cinema parallel in a shopping gallery. A place available to everyone, not just shoppers. Let’s turn customers into viewers and viewers into customers! 6 years later the Mall.cz eshop is acquired by Allegro group and Mall.tv is rebranded to Fameplay.tv.

Content strategy and production ranging from TV series and shows to live broadcasts of the most prestigious cultural and sports events. VOD and data management platform with content data analysis and conversion analysis.

Our ultimate goal was to create an engaging series aimed at raising awareness about the dangers of cyberbullying. The series unravels the story of ninth-grade student Martin Biederman, who becomes a victim of cyberbullying. Martin is killed in a car accident at the beginning of the series, and his father, Petr Biederman, discovers from his Facebook profile that he was blackmailed by someone posing as a girl named Eliška Svobodová.

With the contribution of key experts on cyberbullying and the support of major technological partners, we helped develop a concept and series that captured the attention of juries at several film festivals worldwide. The series was distributed not only on the internet but also in cinemas and on international television channels.