Skip to content

First AI News Anchor

for Frekvence 1

# AI Application

Active Radio, a.s.

2025

F1_Uvodka

We developed the first fully automated virtual AI news anchor for Frekvence 1. The system autonomously collects news from multiple sources, processes them, and delivers broadcasts in both audio and video formats using the voice and likeness of a real presenter.

Fully automated

from news collection to presentation

Audio + video

AI anchor with a real voice

Twice daily

morning and evening summary

AI news

anchor

The virtual anchor processes and presents news from multiple sources in a fully automated pipeline—from content collection and editorial filtering to text-to-speech generation and video rendering with the likeness of a real presenter.

AI news

anchor

The virtual anchor processes and presents news from multiple sources in a fully automated pipeline—from content collection and editorial filtering to text-to-speech generation and video rendering with the likeness of a real presenter.

AI-generated

news website

Alongside the AI anchor, we built a complete news website at frekvence1.cz/zpravy-ai where stories are published automatically. The site features full-text search, category filtering, and responsive design optimized for mobile consumption.

F1_2

Orchestration

AI pipeline

Under the hood, it's an orchestration of multiple AI models in a single pipeline: text processing, audio generation from a cloned voice model, video synthesis with lip-synced presenter footage, and automated publishing across web and social channels.

What our client

client testimonial

AI_moderator_testimonial
F1_lista

“We're seeing growing demand for innovative content. Working with Cognito allowed us to bring listeners an entirely new experience—an AI-powered news service that delivers reliable, timely updates twice a day.”

Lubor Zoufal Head of Digital Products, Active Radio

Jak jsme our approach

From AI model evaluation through orchestration layer development to deployment of a fully automated news production system.

1

AI model evaluation and selection

We evaluated available models for text processing, voice synthesis, and video generation. We designed a pipeline architecture that orchestrates multiple AI models in a single automated workflow.
2

Orchestration layer development

We built a system that connects news collection, processing, prioritization, and summarization with audio and video content generation using a cloned voice model and lip-synced presenter footage.
3

News website

We developed a complete AI-generated news website for automated story publishing.
4

Deployment and operations

System deployed to production, generating morning and evening news summaries without human intervention.
PHP
PHP
Symfony
Symfony
React
React
Docker
Docker
K8s
Kubernetes
Redis
Redis

Technologies technologies

The Frekvence 1 AI anchor runs on PHP and Symfony with a React frontend. The pipeline orchestrates multiple AI models for text processing, voice synthesis, and video generation.

Services and products

services we delivered

Have a project or an idea? Schedule a free consultation and find out how we can help you.

The Cognizance Blog Honest developer insights

Not enough? We have more