AI Agents Deliver Personalized Night Radio Broadcasts, Cutting Costs by Up to 60%
Preparing Nightly Radio Broadcasts Required a Full Team Doing Repetitive Manual Reformatting
Seznam sought to automate repetitive tasks involved in converting content into radio broadcasts, enabling human staff to focus on creating quality material rather than reformatting it. The project also aimed to explore personalisation options such as region-specific newscasts.
Two AI Agents: One Curates Playlists and News, One Writes Scripts with Real Presenter Voices
A specialised AI agent system was implemented where each agent handles specific broadcast production tasks. Agent #1 (Radio Broadcast Producer) generates time-optimised playlists, selects news articles, curates music, and schedules elements. Agent #2 (Radio Host & Script Writer) uses LLMs to write engaging scripts based on news, weather, and content guidelines, then applies text-to-speech technology trained on real presenter voices.
Night Shows Fully Automated, Costs Cut 40–60%, One Person Now Oversees What a Team Once Did
Night programs now run fully automatically without live staff. Operational costs were cut by 40–60% through automated script creation and content selection. One person can now oversee final daytime output versus the previous team requirement. The system opens the door to region-specific personalised content delivery.
More Case Studies
Global Cybersecurity Vendor
ML Pipeline Processes Terabytes of Cybersecurity Data Daily
A global US-based cybersecurity vendor needed to reduce the volume of threats their analysts manually review. Blindspot built a scalable unsupervised ML pipeline for anomaly detection on time-series data, capable of processing terabytes of security events daily.
Expert ML Team Enhances DFLabs' Cybersecurity Product
DFLabs required specialised AI and ML expertise to strengthen their cybersecurity product. Blindspot provided a dedicated expert team who implemented machine learning capabilities directly into the product offering.
Network Monitoring Provider
Real-Time Network Anomaly Detection Deployed Across Client Base
A network performance monitoring provider needed a scalable ML solution deployable across their entire customer base. Blindspot designed a pipeline that processes diverse network indicator data through multiple anomaly detectors, enabling real-time problem identification.
Want to be Our Next Success Story?
Our team of AI engineers and domain experts will work with you to understand your challenge, design a solution, and deploy it to production.