Languages
Rust, Go, TypeScript, Deno, Python, Kotlin, C#
hey, i'm kevin arellano
infrastructure engineer who builds things that scale (and occasionally break in interesting ways)
FedRAMP-High • AWS GovCloud • Zero Trust • 7+ years making backends behave
see what i've builtabout me
I'm Kevin Arellano, an infrastructure engineer who loves turning messy legacy systems into clean, scalable architectures. Spent the last 7+ years modernizing distributed systems, wrangling compliance (FedRAMP-High, HIPAA, PCI DSS), and building cloud-native infrastructure that actually works. Currently working in AWS GovCloud on defense-critical systems, which is as serious as it sounds but also surprisingly fun.
I believe in writing code that's maintainable, systems that are observable, and documentation that doesn't make people cry. Also a big fan of Rust, terrible puns, and reducing deployment times from hours to minutes. Based in Atlanta, GA, I work with teams across the US building secure, scalable backend systems.
things i work with
AWS
GovCloud, Lambda, S3, Kinesis, Glue, SageMaker, DynamoDB, Terraform, Terragrunt
Google Cloud
Cloud Run, GCE, Pub/Sub, Kubernetes
Observability
Prometheus, Grafana, ELK Stack, Structured Logging, Distributed Tracing
Compliance
FedRAMP-High, PCI DSS, HIPAA, Zero Trust Architecture
Architecture
Microservices, Serverless, IaC, CI/CD, GraphQL, Incident Response
stuff i've built
govcloud infrastructure
multi-account AWS GovCloud setup with FedRAMP-High compliance. automated governance, cost optimization, and Zero Trust identity systems for defense workloads
view on github →the 94% faster pipeline
rewrote a daily data pipeline in Rust that went from 14+ hours to 45 minutes. concurrent processing + streaming ingestion = happy data team
coming soon →monolith → microservices
transformed a struggling GraphQL monolith into serverless microservices on GCP Cloud Run. fixed scaling bottlenecks and made deploys actually work
coming soon →hipaa-compliant ml pipeline
built biometric data ingestion on AWS (S3, Kinesis, Glue, SageMaker) handling hundreds of events/sec while keeping lawyers happy
coming soon →observability that works
centralized monitoring with Prometheus & Grafana. structured logging, distributed tracing, and dashboards that actually help debug things
coming soon →24 hours → 60 minutes
modernized a legacy fulfillment platform from monolith to modular system. deployments went from a full day to under an hour
coming soon →let's chat
Email:
kevarellano@proton.me
GitHub:
github.com/kevarellano
LinkedIn:
linkedin.com/in/kevarellano
Location: Atlanta, GA