Case Study: VTS achieves faster ML model productionization with Provectus

VTS productionizes ML models more efficiently, and on a larger scale, accelerating time to market for ML applications

VTS, Inc., a commercial real estate leasing and asset management software and data company, wanted to productionize machine learning models more efficiently and build new models iteratively using AWS services. Although its data scientists could prototype models in notebooks, VTS lacked the AWS and MLOps expertise to move them into production, so it worked with Provectus.

Provectus delivered a template-based ML infrastructure solution, including the VTS ML Commons SDK, Amazon SageMaker job, pipeline, and endpoint templates, plus documentation and training. This helped VTS deploy a leasing deal outcomes model in production, reduce manual work and human error, and speed up future ML delivery; the templates and SDK were delivered within three months and enabled about 3x faster at-scale model delivery.

VTS, Inc.