I run AI platforms in production, under regulated and high-traffic load.

Blaze

Senior SRE · Platform Lead

Lead platform reliability and SRE for compliance-aligned AWS infrastructure serving regulated Brazilian betting brands under high concurrency. I designed the four-layer Terragrunt hierarchy that lets a new brand or environment stand up by copying one directory and editing two files.

EPAM Anywhere

Lead Systems Engineer · Infrastructure & Generative-AI Platforms

Multiple infrastructure engagements at EPAM Anywhere, all on AWS, all Terraform-modeled, all shipping through GitHub Actions with OIDC and zero long-lived keys. The centerpiece is a multi-environment AWS platform for an AI-powered customer-insight product — around 25 services across an LLM gateway (DIAL Core, DIAL Chat, DIAL Themes, DIAL Bedrock Adapter), a roster of domain LLM agents, Prefect-driven document and transcript pipelines, observability, SSO, and shared databases, all running on a single ECS Fargate cluster. Two parallel engagements: an EKS-hosted GenAI tutor for K-12 (DIAL on EKS + Azure OpenAI + MilvusDB with a custom tenancy API) and infrastructure tooling for an energy-sector client (EKS, Amazon Managed observability via Terraform, Kubernetes 1.21 → 1.23 upgrade under live load, Nexus-backed private Python supply chain, blue/green via ExternalDNS). I'm a contributor to the open-source DIAL project that underpins the LLM gateway on both AI platforms.