Michael.VanderValk
Senior Product Leader · People & Agents

Lead. Build. Ship.

Leading teams — human and AI — that build great products.

The next generation of product leadership means directing teams of people and teams of agents toward the same outcome. I've spent my career building both — scaling AI, data, and analytics platforms at Amazon, leading product and UX transformation at Deloitte, and earning an NYU Tech MBA to bridge technical systems and business outcomes. Now I build in public: shipping tools, directing agent teams, and documenting what actually works.

What I Do Now
Lead AI-era product work across apps, agent systems, and workflow design with a focus on practical implementation.
What The Builds Prove
I can move from strategy to shipped product and show the operating model behind the output, not just demos.
Current Focus
Momentum (AI-powered patient intake for PT) and a production-grade AI Personal Operating System that processes daily work context overnight.
4
Apps Live
5
Deployed Agents
Daily
Personal OS Runtime
Michael VanderValk

Headshot generated with StudioAI Photos — an app I built.

Career arc · click
Select a timeline item.
Deloitte
12 years leading product and UX transformation — rebuilding the Audit business for the digital age
12 years
NYU Stern
Tech MBA — bridging business strategy and technical systems
1 year
Amazon
Led AI & data platform initiatives at scale across multiple organizations
4 years
Now
Shipping apps, directing agent teams, building in public
4 apps live · 5 agents deployed
6+
Builder Tools Used
Portfolio Logic
Background -> Forward Focus
Background
Enterprise Transformation Systems AI + Data
Forward Focus
HealthTech FinTech B2C Apps AI Product Mgmt
Portfolio Map
Same operating pattern, different domains
Capability Origin
Systems / Transformation AI Product Mgmt
Forward Focus Lanes
HealthTech FinTech B2C AI PM
Why These Builds Belong Together
The projects look different on the surface, but they are all the same operating pattern: take messy human workflows, design the right AI-human handoffs, and produce something people can actually use in real contexts.
Portfolio View

A collection of AI-first systems built to explore how generative AI reshapes real-world workflows. Developed through coordinated agent teams, each project experiments with eliminating friction, compressing time, and turning ideas into working products fast.

Live
PM Interview Coach
Capability: AI PMAI PM
Voice-enabled drill app for senior PM interview prep. Seven structured drills targeting specific sub-skills — ecosystem mapping, prioritization, root cause analysis — with AI-scored reps benchmarked to L4/L5/L6 levels.
Live
StudioAI Photos
Capability: SystemsB2C
Turns casual selfies into professional LinkedIn headshots using Gemini image generation. Upload up to 5 reference photos, pick attire and background style, download a studio-quality result. Built in an afternoon — because the alternatives cost $40.
In Motion
Momentum
Capability: Systems / TransformationHealthTech
AI-powered patient intake for physical therapy. Patients organize their clinical story before the visit — Momentum synthesizes it into a structured brief that's ready before the provider walks in. Designed for clinics and health systems rethinking the full patient journey.
Live
AI Personal Operating System
Capability: Systems / TransformationAI PM
A CLI-driven agent stack running inside Obsidian. Processes daily notes overnight — extracting tasks, routing them to the right projects, scoring priorities, and generating weekly reflections. Running daily since late 2025.
In Motion
HH Payroll
Capability: SystemsFinTech
Full-stack household payroll app. Handles gross pay, tax withholding, PTO accrual, versioned pay statements, and payment tracking — built to replace a spreadsheet with something actually auditable.
In Motion
unFlakeable
Capability: AI PMB2C
AI agent that wrangles your friend group into committing to a plan — through a live War Room feed with a selectable agent personality. Drill sergeant, hype man, chaos agent — pick your poison.
// Up Next What I'm building toward
Career
BestSelf.ai
Structured AI interview that helps you articulate who you are at your best — for job apps, interviews, or personal clarity.
AI Infra
Agent Workflow Builder
Visual tool for designing multi-agent product workflows. Miro meets LangGraph — for PMs, not engineers.
Productivity
PM Standup Agent
Pulls context from Jira, Slack, and docs to auto-draft your daily standup. Saves 10 minutes, keeps it honest.
Professional Experience Amazon + Deloitte portfolio

I lead the transformation of complex enterprise workflows into scalable, AI-powered platforms. Across Amazon and Deloitte, I’ve partnered with engineering, data science, UX, and executive leaders to embed big data and machine learning into revenue-critical and regulated systems — while helping build the teams and operating models behind them.

I understand both how to prototype AI systems quickly and how to productionize them inside complex, regulated enterprises — aligning data, UX, engineering, and operating models so AI becomes infrastructure, not a demo.

Professional
GenAI Research Assistant
EnterpriseAI + DataSystems
Turns scattered institutional knowledge into instant answers for analytics teams supporting complex advertiser portfolios. Embedded a RAG-based assistant directly into Slack and internal tools so research that took hours now takes minutes — replacing fragmented knowledge systems few people consistently used.
Professional
Insight Automation Platform
TransformationAI + DataEnterprise
Takes advanced ML analyses trapped in Python notebooks and turns them into guided, low-code workflows anyone on the analytics team can run. Scaled bespoke data science into repeatable, production-grade systems that made complex analytics accessible across a global organization.
Professional
Predictive Audience Engine
AI + DataEnterpriseSystems
Combines billions of first- and third-party signals into privacy-safe ML audiences advertisers can actually act on. Shifted targeting from reactive reporting to predictive decision-making, embedding machine learning directly into campaign planning workflows.
Professional
Performance Benchmarking Platform
EnterpriseSystemsAI + Data
Gives advertisers real-time, peer-level performance context instead of static reports. Built a benchmarking layer that surfaces proactive opportunities directly within workflow — replacing slow, fragmented performance analysis.
Professional
Non-Endemic Insight Expansion
TransformationEnterpriseSystems
Created scalable insight delivery for brands that don’t sell on Amazon but advertise there. Defined the roadmap, built the organizational foundations, and productized high-value insights so a previously underserved segment could access structured, data-driven guidance.
Professional
Deloitte Omnia (Global Audit Platform UX)
EnterpriseSystemsTransformation
Turned a fragmented set of audit tools into a cohesive global platform used across multiple services. Unified onboarding, navigation, dashboards, and integrations — making complex regulatory workflows feel like one product instead of many.
Professional
Document Intelligence & ML Automation
EnterpriseAI + DataTransformation
Replaced manual document review in audit engagements with ML-powered extraction and analytics. Integrated machine learning into contract, lease, and financial statement workflows — embedding AI directly into regulated enterprise work.
Professional
Audit Confirmation Platform
EnterpriseTransformation
Enabled audit teams to design custom web-based confirmation forms, send thousands of secure requests, and track responses end-to-end. Automated follow-ups, reduced the risk of misdirected sensitive information, and streamlined processing of completed confirmations — while preserving the integrity of direct third-party verification.

These are the agents running in my personal operating system — real, deployed, and processing my work daily. Each one has a defined job, runs on a schedule or trigger, and outputs something I actually use. This is what "leading agents" looks like in practice.

Personal Operating System Daily Systems
Claude Code · bash · git
Daily Note Processor
Extracts tasks and ideas from freeform daily captures, routes them to the right project or area file, and classifies each as Action (⚡) vs Motion (🔄).
Claude · priority model
Priority Scorer
Ranks all open tasks 0–100 using a multi-factor model: due date, blocking status, manual urgency flags, and goal alignment boost.
Claude · weekly context
Weekly Review Agent
Synthesizes 7 days of daily notes into a structured reflection: task counts by category, Action:Motion ratio, productivity patterns, and next-week recommendations.
Claude · web search · resume
Interview Prep Builder
Combines web-fetched company research, a STAR story bank, and your resume to generate a tailored brief for any interview type in minutes.
Claude · semantic matching
Duplicate Detector
Prevents redundant notes before filing — title matching → keyword search → semantic similarity at 85%+ confidence before anything new is created.
01
Treat Your AI Like an Employee — Because That's What It Is
The intern → junior → senior framework for calibrating how much direction AI needs at each stage of your working relationship.
02
How I Use Obsidian to Give AI Perfect Context
The system that helped another PM ace an HM interview round. Simple to set up, permanently changes how you work with AI.
// Builder Stack Tools I ship with
Claude Code
Agentic coding · full-stack
Cursor
IDE · AI pair programming
Google AI Studio
Prototyping · multimodal AI
Anthropic API
Claude · production AI
Supabase
Realtime · auth · database
Vercel
Deployment · edge functions
Codex
Code generation · CLI agent
// Get in touch

Let's talk product.