Twins AI moved its production AI voice platform from GCP to Azure (www.twinsai.com)

Twins AI has successfully migrated its AI voice platform from Google Cloud Platform (GCP) to Microsoft Azure, leveraging $350,000 in Azure credits offered by Microsoft for Startups. This migration, which took just two weeks to execute, allowed Twins AI to reduce their infrastructure costs significantly while maintaining the same application architecture based on containers and Kubernetes. The move was made seamless for customers, with no disruptions in service, and resulted in improved platform performance. The significance of this transition for the AI/ML community lies in Twins AI's strategy to utilize architectural best practices like GitOps and infrastructure as code, which facilitated a portable design and efficient deployment. The migration capitalizes on Azure's Foundry catalog for model selection, enabling Twins AI to optimize their voice platform with advanced models such as Azure OpenAI while maintaining existing functionalities. This case exemplifies the importance of portability and strategic planning in cloud migration, encouraging startups to adopt modular architectures that enhance flexibility and expedite transitions without extensive recoding.