AI DevOps & Infrastructure
Peoples Energy Analytics: Multi-tenant cloud architecture
"Working with Tech 42 was an exceptional experience. Their team designed and implemented a secure multi-tenant cloud architecture to protect client data, leveraging native and scalable provisioning capabilities to enable seamless account automation and streamlined stack deployment. They also built out dedicated compute environments for model training, ensuring each user’s data was fully isolated within fully segmented accounts.
Communication was clear and consistent throughout and the team showed remarkable patience as we worked through internal resourcing challenges. Tech 42 delivered everything outlined in the project scope, provided thorough documentation, and equipped our team with the knowledge transfer needed to move forward confidently.
We’re extremely satisfied with the outcome and look forward to partnering with Tech 42 again in the future."
Malcolm Blow
Director of Cybersecurity
Project summary
Peoples Energy Analytics needed to scale their machine learning infrastructure beyond local EC2 development. Tech 42 migrated their ML preprocessing and training code to Amazon SageMaker and built a multi-account cloud architecture for secure data isolation. The new platform includes parallelized batch processing for customer data, automated account provisioning via Terraform, and dedicated compute environments for model training. Each customer's data remains fully isolated within segmented AWS accounts. The result is a production-grade system that reduces infrastructure complexity, lowers computational costs through batch processing, and enables scalable machine learning workflows. Tech 42 delivered complete Terraform modules, comprehensive documentation, and knowledge transfer to equip the team to operate and extend the platform independently.
