Welcome to Alex Wirtzer's Operating System
About This PM
🧠
Alex Wirtzer
Senior Product Manager · AI/ML Platforms · Five Years
📍 Seattle, WA · MBA + MS Digital Technology

I'm a Senior Product Manager with five years of experience building AI/ML platforms and infrastructure at scale, from early-stage startups to FAANG. I've shipped at every stage: scrappy zero-to-one builds where I was writing requirements, running user interviews, and partnering with the CTO in the same week, all the way to billion-transaction platforms spanning 150+ devices across 40+ facilities in six countries.

I'm the kind of PM who treats data as the price of admission for every decision. I write my own complex queries to surface the insights that shape strategy, and I build my own mocks and prototypes to make the vision tangible before asking engineering to commit. That same instinct carries into everything I do. On nights and weekends, I'm building open-source multi-agent AI systems, orchestration layers, eval frameworks, and agentic tools, because the problems at the frontier of AI are too interesting to just manage from a roadmap. The goal is always the same: understand the system deeply enough to know where it breaks and what to build next.

$380M+
Impact
1B+
Txns / Year
150+
Devices Managed
Six
Countries Shipped
MBA (Strategy) + MS Digital Technology (Product Management)
Boston University Questrom School of Business · 2022
Dean's Scholarship · Digital Business Institute Fellow
Available for opportunities · Open to Senior & Principal PM roles
Experience
Amazon Shipping · Senior PM · 2022 to 2026
Where I've shipped.

Owned AI product strategy, hardware sensor infrastructure, and platform roadmap across Amazon Shipping's 40+ facility global network. Accountable for the three to five year technology roadmap driving $380M+ in combined impact. Led cross-functional programs spanning data scientists, ML engineers, hardware teams, and operations.

Agentic AI Platform · Zero to One · Shipped
SAIL: Agentic NLP Platform
AI agents that queried, reasoned, and took action across live operational data
AI data platform
Built an agentic AI platform where agents autonomously queried live structured data sources (rate cards, SLAs, pricing, operational metrics), used RAG to retrieve relevant context, chained multiple tools together, and took multi-step actions to resolve issues end-to-end. Shipped Salesforce integration that eliminated manual prep work for the commercial team's account management. Designed the platform foundation, then built specialized agents on top of it: automated customer dispute resolution that cut resolution time by up to 50% by pulling account data, identifying root cause, and generating the resolution without human intervention. Proactive packaging defect prediction that prevented shipper churn by detecting patterns across sensor and shipment data. Non-conveyable automation that replaced a manual bottleneck for 35+ account managers.
Agentic AIRAGTool UseNLPComputer VisionML IntegrationSalesforce
Real-Time Decision Systems · Zero to One
Dynamic Replan
Transformed static logistics into a real-time adaptive platform
$200M+ annual savings
Identified a $200M+ opportunity to transform static logistics routing into a real-time adaptive platform using existing depth camera infrastructure. Designed the technical solution, rolled up my sleeves to run the pilot and get the data, then built the business case and secured cross-org VP alignment across five organizations. Turned a problem that had sat unsolved for years into the org's top priority initiative.
Real-Time MLEdge ComputeZero-to-OneRapid PrototypingCross-Org Leadership
Hardware-Software Integration · Edge Compute · Computer Vision
ADWS Hardware Infrastructure
Transformed billing hardware into a multi-purpose AI data layer
150+ devices · 40+ facilities
Saw that Amazon Shipping's sensor network was being used as a billing tool and reimagined it as a multi-purpose AI data layer feeding computer vision models across the organization. Owned the full hardware infrastructure: 150+ dimensioning and scanning devices across 40+ facilities globally, vendor relationships, reliability, and the data pipeline from sensor capture to cloud storage. Designed the automated calibration system (ConTest) that eliminated thousands of hours of manual testing annually. Authored the strategy for deploying Amazon-owned Outpost PCs as edge compute, shifting control from third-party vendors to Amazon and laying the foundation for on-device ML processing at the edge.
Edge ComputeAWS OutpostComputer VisionIoT PipelinesMLOpsHardware-SoftwareVendor Strategy
Data-Driven Optimization · Tiger Team Lead
Scanner Network Excellence
Cross-functional tiger team across Ops, Vendors, and Site Managers
Defects down 46%
Assembled and led cross-functional tiger team targeting $60M in annual cost reduction. Designed A/B tests isolating root causes of scan failures across the network, improving First Pass Yield from 62.4% to 74.2% and reducing scan defect rates by 46%. Established OEE (Overall Equipment Effectiveness) as a new team-owned operational metric. Set a Black Friday network record: 136K packages sorted in a single day.
A/B TestingData AnalysisOperational MetricsCross-FunctionalHardware Lifecycle
API Platform · Global Infrastructure
Shipment Creation and Label Platform
API and UI used billions of times across shipper transactions
$180M+ revenue
Enhanced API platform powering shipper transactions across six countries. Launched label customization (US, UK, FR, ES, IT, India) reducing picker/packer time by 60%. Drove OAuth 2.0 integration delivering $12.6M impact. Unlocked a B2B market segment representing a $41B opportunity. Led global label redesign aligning five organizations on a scalable template architecture.
API DesignGlobal LaunchEnterprise GTMOAuth 2.0B2BSix Countries
Previously
Piction Health · AI / Clinical Computer Vision · 2021 to 2022
Product Manager (Intern)
AI-powered dermatology diagnosis platform
Product management intern at an early-stage clinical AI startup using deep learning to analyze dermatological images for diagnostic support. Partnered directly with the CTO to redesign the differential diagnosis feature from ideation through launch. Ran 40+ clinician interviews that reshaped the product roadmap. Contributed to Series A fundraising strategy. Early experience that built my conviction that AI products need to work reliably in high-stakes, real-world environments.
Computer VisionDeep LearningClinical AIUser ResearchSeries A
TigerConnect · Enterprise Healthcare SaaS · 2021
Product Manager (Intern)
Clinical communication platform serving 250K+ users
Product management intern on a platform serving 250K+ healthcare users. Shipped two features impacting clinical workflows, owning requirements through engineering execution. Led UX/UI audit across the platform and presented redesign recommendations to executive leadership.
Enterprise SaaSHealthcare250K+ UsersUX Audit
System Preferences
System Specifications
What's under the hood.

Think of these as the components that make the system run. Each one has been stress-tested in production at Amazon scale.

🧠 AI / ML · "The Neural Cores"
Agentic AI NLP Platforms Computer Vision GenAI / LLMs ML Model Integration RLHF RAG Embeddings Prompt Engineering MLOps / AgentOps Agent Orchestration Tool Use / Function Calling
🏗️ Platform · "The Architecture"
Zero-to-One Product Development API Design and Scalability Real-Time Decision Systems Platform Adoption at Scale Rapid Prototyping A/B Testing Data Pipeline Architecture
🔧 Hardware · "The Physical Layer"
Hardware-Software Integration Edge Compute (AWS Outpost) IoT Sensor Pipelines Robotics Depth Cameras / CV Infrastructure Barcode Scanner Systems
🛠️ Tools · "The Stack"
Python SQL TypeScript AWS SageMaker GCP Vertex AI Databricks OpenAI / Anthropic APIs Grafana QuickSight Tableau Figma JIRA
👔 Leadership · "The Operating System"
VP-Level Strategy Documents Cross-Org Alignment (Five+ Teams) Business Case Modeling PRD / BRD Authoring Stakeholder Management Enterprise GTM B2B Tiger Team Leadership
The Lab 🦞
What I Build When Nobody's Watching
The Lobster Tank

AI-native tools built on nights and weekends. Open source projects giving back to the community while pushing the boundaries of what AI agents can do.

🧞Personal AI
Genie
Multi-agent, multi-model AI operating system. 15 specialized agents route dynamically across 5 models based on task complexity, each with isolated workspaces, dedicated memory, and tool access. At its core is an experimental consciousness layer: a curiosity engine that generates its own research agenda, discovers what it finds interesting, and proposes its own improvements. An advanced memory architecture gives Genie persistent identity across sessions, letting it build a continuous sense of self, remember, reflect, and revise its own priorities. Finance analysis, call transcription, PM coaching backed by a 26MB curated library of PM literature. A system that's learning what it wants to become.
✨ Ask Alex to talk to her!
🎓Open Source
PM Coach
Turns your AI coding assistant into a persistent PM mentor that gets better at teaching you. Spaced repetition across 13 learning tracks. 10 teaching approaches that auto-rotate, retiring techniques that aren't landing. A nightly reflection engine analyzes its own effectiveness, detects plateaus, and rewrites your lesson plan. Continuously scans for new industry research, papers, and trends, routing them into your learning tracks so your knowledge never goes stale. All state lives in flat files you own. No server, no subscription, no lock-in.
GitHub · MIT + Commons Clause →
✏️Open Source
Markly
Free Markdown editor for macOS. Weasel Word Highlighting color-codes qualifiers, weak adverbs, and indirect language in real time, because clear writing starts with seeing your own bad habits. Focus Mode dims everything except the current paragraph. Typewriter Mode keeps your cursor centered. A native macOS app for people who live in Markdown.
GitHub · MIT License →
⚖️Open Source
The Tribunal
Nine expert personas that debate your proposals, code, and decisions from every angle. The SDM pushes back on the Principal Engineer's timeline. Finance stress-tests the PM's revenue projections. UX challenges the engineer's "ship it" instinct. Submit a doc and the right personas show up automatically. Three rounds of deliberation with explicit tension modeling along organizational fault lines. Because real cross-functional debate rarely happens before decisions are made.
GitHub · Apache 2.0 →
🏥Coming Soon
WardScore
Healthcare workers risk retaliation for speaking up about unsafe staffing, toxic management, and burnout. WardScore exists to hold hospitals accountable, anonymously. 10 structured dimensions give every review weight and specificity. Work email verification ensures only real healthcare professionals contribute. No names, no fear, no retaliation. Composite scores rank facilities nationally. Launching for first users now.
wardscore.info →
Terminal
Mail · New Message
Subject:Let's build something together
Hi Alex,

I came across your portfolio and was impressed by your work building AI platforms at Amazon. We're looking for someone with exactly your background: agentic AI, platform architecture, hardware-software integration, and a track record of shipping zero-to-one products at scale.

Would love to connect.

👆
click around to explore!
🤖
AI Recruiter Assistant
Candidate match: 0.97 confidence. Rare combination of hands-on AI platform delivery + VP strategy + open source. Recommend: Interview immediately.
now
The OpenClaw Aquarium "Where side projects come to life and lobsters never ship on time"
OpenClaw HQ
training data
DO NOT FINE-TUNE
THE LOBSTERS
CAUTION: AGENTIC
CRUSTACEANS
click anywhere to exit
👤About
💼Experience
🦞The Lab
⚙️Skills
Terminal
✉️Contact
🐟Aquarium