Skip to content

markballew/markballew

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

12 Commits
Β 
Β 

Repository files navigation

Hi πŸ‘‹, I'm Mark Ballew

DevOps Engineer | AI Safety Researcher | Web3 Builder

markballew

ballew

πŸš€ About Me

  • πŸ† ETHDenver 2025 Winner - Won $4,500 with TrustDAI, a decentralized AI trust and verification system
  • πŸ”¬ AI Safety Researcher - Specialized in AI evaluation frameworks and GPU cluster deployment for threat research
  • πŸ›‘οΈ Security & Identity Expert - 15+ years in large-scale automation, DevOps, and security, specializing in decentralized identity and human-agent verification systems
  • 🌱 Currently Building - AI Safety testing framework for Web3 AI agents. Evaluate and monitor agent workflows for unauthorized actions, misalignment, and wallet use
  • πŸŽͺ At EthRome Oct 2025 - Looking for smart contract developers for Web3 AI Agent evaluation and monitoring
  • πŸ’‘ Research Interests - AI agent frameworks, decentralized identity (DID), privacy-preserving data storage, and Web3-AI convergence

πŸŽͺ EthRome Oct 2025

Building: AI Safety testing framework for Web3 agents
Looking for: Solidity developers experienced with test chains and agent interactions
My role: Infrastructure, AI agents, evaluation framework integration

What we're building: Automated testing framework that catches dangerous agent behaviors (unauthorized transactions, misalignment, wallet misuse) in simulated environments before production deployment.

πŸ”§ Current Focus

AI Safety Testing for Web3: Building automated safety evaluation framework for Web3 AI agents using UK AISI's Inspect. Test agent workflows in simulated blockchain environments - catching unauthorized actions, misalignment, and dangerous wallet behaviors before production deployment.

Key Technologies:

  • AI Safety: UK AISI's Inspect framework, Weights & Biases for evaluation tracking
  • Agent Frameworks: LangChain for agent orchestration and testing
  • Blockchain Testing: Foundry, Hardhat, simulated test chains
  • Infrastructure: Kubernetes, OpenTofu/Terraform, Spacelift for CI/CD, AWS/Azure
  • Data Storage: Arweave, Ocean Protocol, IPFS with encryption

πŸ’Ό Experience Highlights

  • METR - Open-sourced Inspect-Action (Inspect extension for Kubernetes GPU clusters) with Terraform and Python, architected EKS deployments for AI threat research
  • UC Berkeley MyShake - Migrated legacy infrastructure to IaC, reduced bugs by 99%, built CI/CD pipelines
  • Okta - Built security DevOps team, implemented PKI audit infrastructure discovering 50%+ idle admin accounts, created automated certificate management systems
  • 15+ years in DevOps, cloud infrastructure, and security automation

πŸ“Š GitHub Stats

GitHub Stats

Top Languages

πŸ† Achievements

  • πŸ“Š 99% Bug Reduction - Infrastructure migration at UC Berkeley
  • πŸ”’ Major Security Improvements - PKI infrastructure and automated certificate management at Okta
  • πŸš€ Open Source Contributor - Co-founder of Chatresearch (700+ member non-profit) and VHSpace.org
  • ⚑ Infrastructure Optimization - Reduced deployment times from months to days through automated CI/CD

πŸ’¬ Let's Connect

πŸ› οΈ Tech Stack

Infrastructure & DevOps

Kubernetes AWS Azure Docker Terraform Ansible

AI & ML

Python PyTorch Weights & Biases LangChain

Blockchain & Web3

Ethereum Solidity Web3.js Foundry

Monitoring & Observability

Datadog Grafana Elasticsearch Splunk


"Building safety testing infrastructure for the AI agent era"

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published