This summer marked a significant milestone in the evolution of PROVYS Sphere, as we introduced a preview of our embedded AI assistant within the web client. What began as a straightforward enhancement—automating translations and generating text metadata—has quickly unfolded into a transformative leap in how broadcasters interact with data, content, and workflows.
From routine tasks to smart creativity
Imagine generating an engaging EPG synopsis with just a click. The assistant takes basic programme data—for example, metadata imported from Opta for a live sports event—and crafts rich descriptions, automatically translating them in to multiple languages. It can even adjust tone and style, making the same content suitable for diverse platforms such as linear schedules or dynamic VOD catalogs. With saved and reusable prompts, broadcasters can ensure consistency across teams and maintain alignment with their brand voice and cultural standards.
Discovering a universal tool
During development, it became clear that this assistant could do far more than text generation. PROVYS users can now combine multiple metadata fields into a single LLM-driven request, enabling complex validations and feedback loops. For example, an editor can feed an entire programme record into the assistant alongside a company-style prompt and instantly receive feedback on metadata quality or completeness. Everything without writing a single line of code, just by prompt engineering.
Similarly, assistant can be used to evaluate larger structures, such as transmission schedules, for coherence, flow, or potential conflicts. This contextual overview helps teams see beyond individual items, allowing for a better understanding of the broadcast lineup.
Looking ahead: AI agents and data intelligence
The roadmap goes even further. PROVYS plans to introduce background AI agents that will monitor schedules and highlight operational priorities—summarizing what requires attention before the day begins. In time, these agents could even invert traditional workflows: generating draft metadata in the background, validating them automatically, and presenting refined results for human review and approval—combining automation with accountability.
A major breakthrough comes with the use of AI-driven analytics on PROVYS data. Through our standard transformation pipelines, extensive datasets — including programme libraries, schedules, and transmission histories — can be exported to a data warehouse and enhanced with vector search technology.
This innovation enables users to ask natural-language questions directly about their data, receiving instant, meaningful answers — without touching spreadsheets, reports, or SQL. It represents a new, intuitive way to interact with PROVYS, ideal for ad-hoc analysis, rapid decision-making, and uncovering hidden insights. We see this as a transformational step in how broadcasters access and understand their operational intelligence — and it deserves a story of its own.
Join the AI-driven future of broadcasting
We are proud of what PROVYS Sphere can already do with AI—and this is just the beginning. We invite our users and partners to explore these new possibilities with us and to share their own AI success stories. Together, we can redefine the way broadcasting operates in the age of intelligent automation. Contact us to discover more.
