Michael Tuszynski

AI Engineering Leader · Platform Builder · Former CTO & AWS Solutions Architect · 25 Years Building at Scale

Profile

I build production AI systems that engineering and revenue teams actually use — not demos, not strategy decks. My current work codifies a Fortune-1000 professional services org's full pursuit lifecycle — discovery, qualification, proposal, SOW, delivery handoff — as an internal AI platform: 11 plugins, 60+ versioned slash commands, ~51K LOC, orchestrating Salesforce, Microsoft 365, OneDrive, Anthropic Claude, and AWS Bedrock behind a single operator surface.

I get to do this credibly because I've spent 25 years in the operations layer underneath: ran a 20-person engineering org as CTO of a streaming platform serving millions, spent six years as a Senior Solutions Architect at AWS migrating the workloads everyone said couldn't be moved, and co-authored the ECS Workshop that has trained thousands of engineers. AI engineering at scale isn't a different sport — it's the same evidence-driven, observable, cost-aware platform engineering, applied to a new substrate.

What I'm hired for: turning hand-wavy AI initiatives into versioned, auditable platforms with real guardrails, real evaluations, and real adoption.

Selected AI Engineering Highlights

Internal AI presales platform (Presidio, 2025–2026) — Architected and shipped a Claude Code plugin marketplace modeling the full pursuit lifecycle. 11 plugins, 60+ slash commands, ~51K LOC. Compresses hours of manual artifact production (call prep, qualification, proposals, SOWs, RAP submissions, delivery handoffs) into command-driven workflows with enforced evidence citation, prohibited-term scrubbing, and Presidio-claim attribution.
Platform-level AI engineering primitives — Built a cross-platform OneDrive path resolver, prompt-contract framework with versioned input/output schemas, evidence-citation and confidence-disclosure guardrails (High/Medium/Low + dated source URLs), and a session-start hook that hot-loads operational knowledge from a Teams-synced channel — institutional memory becomes ambient without a plugin release.
Agentic workflow design — Codified a 57-question DSAE qualification framework, RAP readiness pipeline, SOW redline package generator, staffing/pricing workbook auto-fill with template-drift detection, and cross-source company intelligence with confidence-rated outputs.
Multi-provider AI infrastructure — Anthropic Claude (primary) with AWS Bedrock fallback for users without enterprise access; prompt caching for shared guardrails and knowledge base; per-command rule contracts; provider switching via env-var toggle.
AI advisory at the C-suite — Designed modern application architectures incorporating LLMs and agentic AI for enterprise clients across Financial Services, Healthcare, and Manufacturing — bridging the gap between AI hype and production-ready implementation.

Professional Experience

Principal Architect | Presidio

July 2025 - Present
San Francisco Bay Area · Remote

Architected and shipped Presidio's internal AI presales platform — a Claude Code plugin marketplace that models the full pursuit lifecycle as versioned, auditable workflows (see Selected AI Engineering Highlights above for scope and primitives)
Designed end-to-end AI-augmented solutions and roadmaps for C-suite stakeholders across Financial Services, Healthcare, and Manufacturing — including agentic AI implementations, LLM integration patterns, and evaluation strategies
Lead pre-sales technical strategy for AWS application modernization, conducting customer discovery workshops and defining target architectures for enterprise clients
Design containerization (EKS), serverless (Lambda), and API-first approaches for legacy system modernization
Serve as technical bridge between customers, sales, and delivery — translating business problems into viable solution architectures backed by working reference implementations

Independent Consultant | MPT Solutions

April 2024 - June 2025
San Francisco Bay Area · Remote

Built CI/CD pipelines from scratch for multiple clients, cutting deployment cycles from weekly to daily with integrated security scanning
Led end-to-end development of a cloud-native platform, reducing client time-to-market by 50%
Provided fractional architecture advisory — including build-vs-buy decisions on AI/LLM tooling and helping clients avoid over-engineering early-stage AI features
Implemented security and compliance frameworks preparing client environments for enterprise readiness

Senior Solutions Architect | Amazon Web Services (AWS)

December 2017 - March 2024

Spent six years as a customer-facing SA helping enterprises design and implement cloud-native architectures — primarily container-based (ECS, EKS, Fargate) and serverless
Co-authored the ECS Workshop, one of AWS's most-used container training resources — a platform-engineering project that has trained thousands of engineers globally
Led complex migration projects from discovery through production cutover, specializing in the applications that "couldn't be moved"
Drove consistent cloud cost optimization exceeding 30% for enterprise accounts; featured speaker on AWS re:Think Podcast on right-sizing
Conducted deep-dive technical workshops and architecture reviews for engineering teams across industries

Chief Technology Officer | Fandor

January 2014 - November 2017

Ran the full technical organization — engineering, product, and operations — for a streaming platform serving millions of users. 20+ person team, $4MM+ annual budget.
Architected cloud-native video processing pipeline handling 10,000+ titles with 99.98% uptime while processing petabytes of content
Led the monolith-to-microservices migration that enabled 100% YoY user growth without proportional infrastructure cost increases
Built the AVOD business from scratch, growing ad impressions 200% month-over-month
Established DevOps practices that improved deployment frequency by 10x

Director Technical Services | PointRoll

January 2006 - December 2014

Led Ad Server Engineering and Platform Engineering teams for a digital advertising technology company
Optimized system architecture reducing infrastructure costs by 35% while improving reliability
Implemented disaster recovery program that cut downtime by 50%

US Director of Technology | Flashtalking, Inc.

January 2012 - December 2013

Led a 10-person technology team through rapid scaling, achieving 30% system efficiency improvement
Developed strategic technology roadmap that contributed to 15% revenue growth

Earlier Positions

Web Developer | Gartmore Global Investments (2004 - 2006)
Systems Analyst | Towers Watson (2001 - 2004)

Technical Expertise

AI & Agent Engineering

Anthropic Claude API (Sonnet, Opus, Haiku)
Claude Code plugin development; agent orchestration
Prompt-contract design; structured outputs
LLM evaluation, guardrails & evidence citation
Prompt caching; multi-provider routing & fallback
MCP (Model Context Protocol); AWS Bedrock

Cloud & Infrastructure

AWS (deep): EC2, ECS, EKS, Lambda, S3, CloudFront, RDS, DynamoDB, ElastiCache
Infrastructure as Code: Terraform, CloudFormation, CDK
Containerization: Docker, Kubernetes
CI/CD: GitHub Actions, Jenkins, AWS CodePipeline
Observability: CloudWatch, Datadog, New Relic

Languages & Frameworks

Primary: TypeScript, Python, Go
Proficient: Java, Ruby, C#, Rust
Frontend: React, Vue.js
Backend: Node.js, Django, Rails, .NET Core
Data: PostgreSQL, MongoDB, Redis, Elasticsearch

Architecture & Practices

Platform Engineering
Microservices & Event-Driven Architecture
Domain-Driven Design
DevOps & SRE
Salesforce + Microsoft 365 (Graph API) integration
Agile/Scrum at Scale

Personal AI Lab — NEXUS

Multi-machine personal R&D workspace where I prototype the AI platform patterns that show up in client work. NAS + Mac-mini + laptop topology with ~7 production services (TypeScript/Node), 18+ LaunchAgents, custom Slack bot with auto-discovered slash commands, semantic search over conversation history (Ollama embeddings), self-hosted dashboards behind Cloudflare Tunnels, and integrations with Anthropic Claude, Plaid, Microsoft 365, Synology DSM, GitHub, and Ghost CMS. Provides the test bed for prompt-contract patterns, agent orchestration, hot-loaded knowledge, and evaluation harnesses I later bring to enterprise engagements.

Certifications

Harvard Data Science Review - Agentic AI Intensive (December 2025)
UC Berkeley Executive Education - Professional Certificate in Machine Learning and Artificial Intelligence
AWS Certified DevOps Engineer – Professional
AWS Certified Solutions Architect – Professional
AWS Certified Developer Associate
AWS Certified SysOps Administrator – Associate

Education

Saint Joseph's University - Erivan K. Haub School of Business
BS, Management Information Systems