russell.ballestrini.net/content/pages/cover-letters/2025-11-29-runpod-principal-software-engineer-storage-systems-russell-ballestrini.rst

**Russell Ballestrini**

russell@unturf.com | 860-333-3047 | https://russell.ballestrini.net

November 29, 2025

**RunPod Inc.**

**Re: Principal Software Engineer, Storage Systems**

To the RunPod Hiring Team:

Top 4 reasons my experience aligns with architecting scalable cloud storage for AI infrastructure:

**1. Distributed systems & cloud infrastructure at scale**: At Remind (2017-2020), managed AWS infrastructure for 60+ microservices supporting 30M users during COVID-19 lockdowns when distance learning demand exploded. Architected scalable infrastructure using ECS, EC2, RDS, DynamoDB, & S3, ensuring critical services stayed online during unprecedented traffic spikes. Optimized cloud costs while maintaining performance & reliability for mission-critical systems. Built https://unsandbox.com, an anonymous code execution API supporting 42+ languages with containerized isolation, artifact generation (binaries, WASM), & async execution, demonstrating expertise in multi-tenant compute isolation similar to RunPod's GPU workload challenges. This experience directly translates to RunPod's storage challenges of designing scalable architectures that handle AI/ML workloads efficiently while managing costs & ensuring data durability across distributed infrastructure.

**2. Deep Linux, systems programming & containerization**: 15+ years working with GNU/Linux (Ubuntu, Debian, Alpine), building low-level infrastructure tools & debugging kernel-level issues. Proficient in C, Go, Rust, Python, Bash, the exact stack required for storage systems development. Built containerized execution environments for https://unsandbox.com handling 42 programming languages with secure isolation & artifact management. Architected inference infrastructure at https://uncloseai.com handling GPU workloads for streaming text, voice, & TTS models. This systems-level expertise enables me to design & implement storage backends that maximize performance for GPU compute workloads while ensuring data integrity, security, & proper isolation in multi-tenant environments.

**3. Security, compliance & disaster recovery**: At Cofense (2020-2022), spearheaded SOC2 & FedRAMP compliance initiatives, managing AWS account controls, vulnerability scanning (ECR), & security automation. Built disaster recovery strategies, backup solutions, & access management frameworks using Okta & GitLab. Implemented security best practices across cloud infrastructure ensuring data protection & regulatory compliance. Designed secure multi-tenant isolation for https://unsandbox.com's containerized execution environment. RunPod's storage systems require robust security isolation for multitenant workloads & compliance with SOX, HIPAA, & GDPR. My experience building secure, compliant infrastructure ensures I can architect storage solutions that protect customer data while meeting regulatory requirements.

**4. Performance optimization, observability & principal-level ownership**: Introduced Datadog at Remind & Coursemojo with container/Fargate agents for APM, logging, & custom metrics, building dashboards that exposed performance bottlenecks in distributed systems. Optimized database queries, tuned infrastructure configurations, & reduced latency across services handling millions of requests. Monitored storage performance metrics (IOPS, throughput, latency) to ensure SLA compliance. Bootstrapped multiple production SaaS platforms (https://makepostsell.com, https://remarkbox.com, https://uncloseai.com, https://unsandbox.com) demonstrating end-to-end ownership from architecture design through implementation & operations. Authored "uncloseai. Machine Learning Reference Guide to Inference Clients" covering streaming techniques & API design across 42 languages. RunPod's storage team needs principal-level engineers who can drive architectural decisions, monitor performance at scale, mentor teammates, & build systems that grow with the company's AI infrastructure vision.

The opportunity to design storage infrastructure that powers the next generation of AI innovation combines deep technical challenge with meaningful impact. RunPod's mission to democratize GPU access resonates strongly. I've spent my career building infrastructure that empowers users (educators reaching students, creators selling products, developers accessing ML models). I'm excited to bring my distributed systems expertise, security-first mindset, & passion for scalable architecture to help RunPod build storage systems that unlock AI's true potential.

Sincerely,

Russell Ballestrini