Toby Sanchez
AI Developer · MLOps & Platform Engineer
Summary
Engineer with a career spanning silicon to systems to AI — from VLSI design on Intel-compatible microprocessors and Nvidia's first Xbox GPU, through enterprise cloud platform engineering at Microsoft, to active development of AI-integrated production systems. Brings foundational depth in the mathematical and computational concepts underlying modern machine learning alongside hands-on experience deploying and integrating large language models. Equally effective as an individual contributor, technical lead, or hands-on architect.
Skills
AI/ML: LLM application development · agent orchestration · RAG system design · Anthropic Claude · OpenAI API · MLOps · genetic algorithms · Markov chain Monte Carlo · stochastic optimization · time-series statistical analysis · large-scale data conditioning
DevOps/Platform: Kubernetes · Terraform · Ansible · Packer · Vagrant · Docker · Helm · CI/CD pipeline development · immutable infrastructure · hybrid multi-cloud (Azure · AWS · GCP) · GitOps · FPGA synthesis pipelines
Languages: Python · Go · C · C++ · Java · Perl · Ruby · C# · PowerShell · Bash · Lisp · Verilog · VHDL · Objective-C
Data/Databases: Oracle · MySQL · SQLite · NoSQL · ETL · data warehousing · ORM · LLVM/JVM profiling
Electrical Engineering: VLSI CAD automation · GPU shader modeling · formal verification · embedded/microcontroller design · IoT · JTAG · Algorithmic Test Pattern Generation · power rail analysis
Compliance/Security: HIPAA SME · PII architecture review · secrets management · enterprise-scale regulated data systems
Experience
AI Developer
Freelance / Self-Employed — January 2024 – Present
- Training and deploying local large language models and developing Go application suites for clients migrating away from legacy macOS environments
- System automation and cloud infrastructure development for macOS on Apple Silicon and Intel; applying domain expertise to produce high-quality English-language training data for LLM-driven system automation and administration workflows
- Robotic process automation on macOS leveraging LLM-driven decision layers
- Advising on and implementing LLM-driven solutions for firms of all sizes — drawing on the same migration and modernization methodology applied to Fortune 20 clients during the Microsoft Azure engagement
- Active development of local LLM-integrated media and data management systems on commodity hardware
Freelance VLSI Platform Engineer
Series B Confidential Startup + Related Clients — Jul 2021 – Present
- Infrastructure and VLSI CAD software tooling in the cloud for a stealth silicon IP startup
- Vagrant · Packer · Terraform · Vault · Google Cloud · AWS · Kubernetes · Ansible · Algorithmic Test Pattern Generation · JTAG board-level test tooling · GitOps-based FPGA synthesis pipelines on commodity IaaS
Senior Consultant — Azure Cloud and AI
Microsoft — February 2019 – May 2021
- Developed standardized immutable DevOps patterns across PowerShell, Bash, Terraform, Ansible, Packer, Docker, Vagrant, and Linux; built client-specific CI/CD toolchains in Python and Perl
- Led teams and trained Microsoft Solutions Architects across the company on Azure migration using Kubernetes, Helm, and Terraform
- Clientele: Government of Canada · Wells Fargo · AT&T · The Cleveland Clinic Foundation · USDA Food Safety Inspection Service; Cleveland Clinic engagement included production architecture evaluation for PII compliance, resulting in hands-on HIPAA SME expertise
- Led and trained partner teams from IBM, Tech Mahindra, Wipro, and TCS
- Delivered Ansible playbook development, containerized Java stack performance tuning, Terraform + Azure DevOps core IP integration, application migration assessments, and hybrid multi-cloud IaaS/PaaS/SaaS enablement
DevOps Lead
Karma International / LawLinq.com — Nov 2017 – May 2018
- Responsible for investor-facing application stacks following the loss of the primary developer
- Secrets vault management, SSL certificate lifecycle, remote Ubuntu Linux administration
- PoC planning and migration design for full-stack Django/MySQL applications moving to containers
Principal Engineer / Founder
Yaquazi TechWorks — Phoenix, AZ — 2002 – Jan 2019
- Full-stack B2B software development; Linux/Windows/macOS workflow automation for small businesses
- Analog/digital circuit design, CAD automation, solar/wind energy architecture, electromechanical sensor and control systems
Systems Analyst
Apollo Education Group / University of Phoenix — Phoenix, AZ — 2007 – 2010
- Primary support analyst for 3-tier applications serving 5,000+ staff and millions of students; Java AWT and Spring API demographics apps on Oracle and SQL Server
- Trusted with sole access to live Oracle PII database covering all students, leads, and employees across Apollo and University of Phoenix; regularly audited and background-checked
- Developed genetic algorithms to generate and evolve regular expressions for PII/HIPAA-suspect data discovery across a 100-billion-record student demographics database — enabling remediation of compliance exposure at a scale not addressable by hand-authored patterns
- Rapid delivery tooling in C#/.NET, Java AWT, PowerShell, Perl, Python, and Oracle PL/SQL
Senior VLSI CAD Engineer
Nvidia Corporation — Austin, TX — 2000 – 2001
- Member of the photomask tape-out team for the first Microsoft Xbox GPU and its retail counterpart
- Developed automatic buffer insertion flows in Perl and C++ using iterative ML algorithms to optimize cross-chip signal propagation
- GPU architectural test-benches in Verilog, VHDL, C, and C++; formal verification of shader unit functionality in C
Senior VLSI CAD Engineer
VIA Technologies (Centaur) — Austin, TX — 1999 – 2000
- Integrated routing and placement workflows with IC Craftsman and Cadence Virtuoso in Perl and Lisp
- Developed genetic algorithms to improve chip-wide power distribution across 3 Intel-compatible Centaur microprocessor cores — stochastic worst-case test vector generation using Markov chains, optimized via genetic algorithms; visualized via Perl/ImageMagick pipeline
- Automated current-density and voltage-drop analysis using Railmill, Perl, and Cadence Virtuoso
Early Career
VLSI Electrical Engineer/Scientist — Hewlett-Packard, Fort Collins, CO (1998–1999)
IA-64 enterprise CPU timing validation; test vector generation automation combining HP/Intel tools with third-party Algorithmic Test Pattern Generation; automated design process flows in make, bash, tcsh, and Perl.
Component Design Engineer — Intel Corporation, Chandler, AZ (1994–1998)
Wafer and device-level characterization of microprocessor products; CPU cache timing simulations for Pentium III; test hole resolution for 486DX4; test socket temperature control module for the first mobile Pentium.
Education
Bachelor of Science, Electrical Engineering — Northern Arizona University, Flagstaff, AZ