~/sam-bretzcareertagsaboutrsslinkedin (opens in a new tab)x (opens in a new tab)

Career

Staff engineer building production agentic systems. Eight years across platform, data and distributed systems teams, from surgical data pipelines to enterprise identity graphs to AI agents.

Experience

  1. Sep 2025 to now
    Staff Engineer, Agentic Systems
    AG2 (formerly AutoGen)

    Building the core execution systems behind AG2's first commercial agent platform: task loops, scheduling and triggers, tool calling, GraphRAG, memory and multi-model routing. Running a live agentic system for a biotech platform company, leading two enterprise pilots with Fortune 500 telecoms, and building the evaluation tooling that gates weekly releases.

  2. Jan 2025 to Sep 2025
    Team Lead, Automation and Quality
    Veza

    Founded a new team at the CTO's request to build system-level test automation for enterprise scale. Took the demo environment from fully manual to automated and deployed in under an hour, and wrote the scale testing procedures behind Veza's largest deals.

  3. Jan 2023 to Jan 2025
    Senior Software Engineer, Platform
    Veza

    Moved the data pipeline's extract stage from batch ETL to streaming. Cut the Okta integration's runtime by 96.6%, sped up graph ingestion 3x for every integration, and cut Snowflake data costs by 80% while holding freshness SLAs.

  4. Sep 2022 to Nov 2022
    Founding Engineer
    Sieve

    Moved the product from Cloud Functions to Kubernetes on GCP, made video processing 4x faster, and set up metrics, logging and alerting in every environment.

  5. Jun 2021 to Sep 2022
    Senior Software Engineer, Data Platform and Agent
    Automox

    Broke core monolith services into Go microservices and moved from RabbitMQ to Kafka. Built an event platform handling 10,000 to 25,000 events per second, and led five engineers on cross-service tracing that the whole company adopted.

  6. Oct 2018 to Jun 2021
    Software Engineer, Operative Intelligence
    Intuitive Surgical

    Built AWS data pipelines that turned surgical data from hospitals worldwide into datasets serving terabytes of video and time series data daily, and put a TensorFlow model into production that identified the procedure being performed from endoscopic video alone.

Education

  1. 2014 to 2018
    B.S. Computer Science
    Georgia Institute of Technology

    Concentrations in Artificial Intelligence and Information Systems.

Elsewhere

linkedin, x

Tools

Languages
TypeScript, Python, Go, SQL, JavaScript, C, C++, Java
Infra and data
Docker, Kubernetes, Terraform, AWS, GCP, Kafka, RabbitMQ
Frontend
React