This article describes how Avahi Technologies helped MasterWorks migrate a machine learning application from a SaaS provider to AWS, achieving significant cost savings and performance improvements.

• MasterWorks migrated 200 ML models used by 35 clients to Amazon SageMaker in 8 weeks
• Reduced hosting and licensing costs by approximately 75% through AWS on-demand pricing model
• Implemented serverless inference pipeline using AWS Lambda and AWS Glue for cost optimization
• Accelerated model training and inference processes compared to previous SaaS platform
• Achieved zero post-production issues after going live
• Avahi provided detailed documentation enabling internal team cross-training and model development
• MasterWorks now invests cost savings in additional ML initiatives and model improvements

The migration demonstrates how AWS partnerships can deliver ML platform modernization with substantial cost reductions while maintaining service quality and enabling faster innovation.