Grey Newell

ML infrastructure engineer building evaluation, inference, and observability systems for production AI. Creator of the MIST stack, an open-source eval and inference platform in Go. Founding Engineer at Supermodel, building code analysis tooling for AI agents. Previously SDE and Solutions Architect at AWS. Pursuing MS in Computer Science (Machine Learning) at Georgia Tech.

Experience

Founding Engineer

2025 - Present

Supermodel Tools

Building code analysis tooling for AI coding agents: call graph generation, impact analysis, dead code detection
Designed and implemented MCP server enabling sub-second symbol lookups and cross-subsystem analysis via pre-computed code graphs
Built evaluation pipeline using SWE-bench to measure agent performance with and without code analysis tools
Tech: Python, TypeScript, Neo4j, MCP, static analysis, graph databases

Creator, MIST Stack (Open Source)

2025 - Present

Independent

Created MatchSpec (evaluation framework), InferMux (inference routing, 89+ GitHub stars), TokenTrace (token observability), and SchemaFlux (structured data compiler)
All written in Go with zero external dependencies; supports distributed tracing (W3C), circuit breaking, and rate limiting
InferMux routes inference requests across LLM providers with load balancing, failover, and cost optimization
Published research paper benchmarking MCP servers on 500 SWE-bench Verified tasks (DOI: 10.5281/zenodo.18627369)

Senior Solutions Architect

2024 - 2025

Amazon Web Services (AWS)

Advised Amazon Finance on event-driven architecture processing 86M+ daily events for financial monitoring
Published on AWS Architecture Blog and AWS Training & Certification blog
Completed all 12 AWS certifications. Contributed infrastructure as code to aws-samples open source projects

Solutions Architect

2022 - 2023

Amazon Web Services (AWS)

Co-authored "5 tips for AWS Certification exams" on the AWS Training & Certification blog
Earned 6 AWS certifications. Technical advisor for enterprise cloud architectures

Software Development Engineer

2021 - 2022

Amazon

Built distributed event-driven services

Engineering Lead

2019 - 2021

CareerBuilder

Led engineering team building job matching and recommendation systems
Implemented data pipelines and search infrastructure serving millions of users

Education

Georgia Institute of Technology

2025 - 2028

Master of Science in Computer Science · Machine Learning Specialization

Research & Publications

Technical Skills

Python Go TypeScript Model Evaluation Inference Optimization LLM Infrastructure Distributed Systems Docker AWS (12x Certified) Neo4j MLOps RAG Prompt Engineering Event-Driven Architecture MCP