Eli Stark

SUMMARY

Infrastructure engineer with 15+ years in SRE and platform engineering. Led teams scaling systems for millions of users at 300+ Gbps throughput across a fleet of 3,000 bare-metal and cloud servers. Currently building agentic AI systems that automate vulnerability remediation, CI/CD troubleshooting, code review, and dependency management using LLMs and multi-agent architectures.

EXPERIENCE

Carta 6 yrs 10 mos total

Developer Experience & Agentic AI Engineer

Aug 2022 – Jun 2026 // San Francisco Bay Area

Built CI/CD troubleshooting agent that monitors pipeline failures and iteratively applies fixes through feedback loops
Developed AI agents automating major Python and Django upgrades across 1M+ LoC codebases, handling dependency conflicts and breaking changes
Created LangChain multi-agent system scanning repos for vulnerable code and generating security patches organization-wide
Built continuous delivery service (Python, ArgoCD, Slack, CircleCI) standardizing declarative CI/CD and improving deployment success rates
Contributed to a Claude Code-based AI coding harness; maintained its internal marketplace; developed and evaluated agents, skills, and deterministically orchestrated workflows
Contributed to a multi-agent pull request reviewer combining static analysis, repository history, and dependency intelligence to detect high-impact defects before merge
Developed onboarding automation reducing time-to-first-commit from 2-3 days to ~1 hour via automated environment provisioning
Created containerized developer environment using devcontainers with secrets management, linters, and IDE plugins

Site Reliability Engineer

Sep 2019 – Aug 2022 // Palo Alto, CA

Built EBS volume pre-population service improving sequential read performance 25x and reducing dev environment deploy time 65%
Developed ChatOps platform automating onboarding/offboarding across internal and third-party services, saving 30-60 min/day
Reduced Kubernetes costs 50% in lower environments by optimizing deployment lifecycle and purging stale resources
Collaborated on migration of production Kubernetes clusters from kops to AWS EKS, reducing operational overhead