Tailshift: AWS migration assessment
“Tech 42’s AWS assessment helped us gain a clear understanding of how to scale our generative AI platform more efficiently. Their team conducted a thorough evaluation of our summarization and policy-claim matching workflows and presented a well-structured AWS-native architecture, complete with performance benchmarks and cost modeling.While we have decided to hold off on full migration until we reach 10x our current scale to maximize funding and impact, the insights and future-state roadmap provided by Tech 42 have positioned us to move quickly when the time is right. Their strategic guidance and technical depth made this a high-value engagement for our team.”
Armin Pourshafeie
Co-Founder and CTO
Project summary
Tailshift's insurance claims analysis platform processes high-volume summarization and policy-claim matching workflows on OpenAI. The team needed to understand how to scale cost-effectively while handling burst workloads up to 5,000 claims per transaction. Tech 42 conducted a comprehensive assessment: reviewed current architecture, benchmarked 5-10 AWS Bedrock models against the existing OpenAI implementation, designed a future-state AWS-native architecture, and modeled cost optimization with AWS funding recommendations. The engagement delivered a clear performance baseline, model selection guidance, and a strategic roadmap for migration. Tailshift now has the technical foundation and business case to move to Bedrock when scale reaches the target threshold, maximizing both operational efficiency and AWS funding impact.
