Case Study: Radient achieves faster, lower-cost agent routing with Inception Mercury

Instant Decisions for Faster and Cheaper Agentic Work

Radient, an AI research and development company, faced a challenge with the latency of the model routing system within their product, Radient Automatic. Slow router decisions were degrading the user experience for their small business clients, as delays would compound over the course of a complex AI agent's workflow. They needed a fast and accurate solution to classify tasks and their difficulty in real-time.

By implementing Inception's Mercury diffusion model, Radient achieved sub-second routing decisions. Mercury’s parallel-generation architecture provided the required speed for structured outputs, running 5 to 10 times faster than other models. This solution from Inception enabled up to 70% cost savings for end-users by efficiently routing tasks to the most appropriate model and made complex agentic assistants feel instantly responsive.

Radient

Damian Tran

Founder