Watch a leaf fall. Nothing about it is forced.
Great work, in any discipline, starts the same way.
Experiment. Observe. Let the evidence lead.
The best solutions aren't drawn up in one pass. They're grown.
Human-First AI Transformation
Founded in First Principles
Watch a leaf fall. Nothing about it is forced.
Great work, in any discipline, starts the same way.
The best solutions aren't drawn up in one pass. They're grown.
With AI, what stands above ground can now go up in weeks.
What holds it up is harder to see.
It starts below the surface, with the problem worth solving.
And it only stands if it's rooted in the people you mean to empower.
The current never stops. Never in one place, never just one thing.
Principles are what carry people through it, elegantly and intentionally.
Experience
Most of my career has been at Microsoft Security. Since 2025 I've moved from landing strategy with sellers and architects to designing and building the AI tools leadership now uses to see the business and act on it.
I built the program and its AI tooling end to end: monthly calls with 20 business units, AI recaps within a day, and owned actions that unblock customer deals and escalate critical product blockers to engineering, reported back to leadership and sellers.
The original criteria matched no one. I designed the method and built the AI model end to end, from architects' own learning and delivery records to scoring, cohorts and manager review.
I designed the paths and the readiness data behind them, then landed them with leadership and ~780 architects, with reporting down to each person.
I built AI agents that pair each message with live readiness data and insights for that person's team, which lifted training completion in the regions that were behind.
Microsoft Global Customer Success Award
FY26 Q4 · The Accelerators team awardI lead AI tooling and business insights for the Security business: the systems that turn scattered data, conversations and activity across the business into insight and owned action.
Ran enablement for Security's solution architects as one system, with data, strategy and training working together.
University of Washington, Foster School of Business
BA, Business Administration (Marketing) · 2017 – 2019Bellevue College
Associate's in Business Administration · 2015 – 2017Selected work
Each one started with a business question leadership couldn't answer yet. Pick one to read the full story.
Move to aim · click to loose

About
I'm a business strategist at Microsoft Security. Most of my work comes back to two questions: what is the business really telling us, and what should we do about it? I design and build the AI systems that answer them, then stay with them until the people they're for are actually using them.
I'm not an engineer, and that's by choice. AI does the engineering; I own the problem, the design and the landing. Everything I build starts from evidence that already exists, keeps people making the calls, and closes the loop with the people it affects.
Off the clock I'm a published nature and backpacking photographer, which is where most of the ideas on this page began.
Short film
Five years inside Microsoft, the systems I've built, and why AI adoption should end in human flourishing.
Contact
If something here resonated, I'd love to hear from you.
01Case study · Field Insights Platform
Asks reached sellers and architects from every direction, and nobody could see what happened next. Leaders saw the numbers, but the story behind them lived in scattered chats, emails and 1:1s, and sellers rarely heard what became of their asks.
One monthly rhythm that pulls in calls, chats, email, trackers and the weekly numbers from every business unit: Listen, Understand, Act, Close the Loop. AI connects the signals, drafts each recap within about a day, cites every claim to the exact recording moment and has a second model fact-check it. People stay in charge of every decision and every message.
From scattered asks to a closed loop, made the official system by leadership.
Insights a month, backed by thousands of citations to the source
Owned actions a month, each with a named owner
Operating units where customer deals were unblocked
From signal to a cited, fact-checked recap








Every UI visual uses demo data; figures in the copy describe the program itself.
Build systems that run without their builder. The platform runs in the cloud, not on my PC; only the monthly recap step stays AI-assisted by design.
02Case study · Capability Intelligence Model
Leaders had to choose experts for flagship AI-security work, and the criteria as written matched no one: 0 of 895 architects, because the certifications they named didn't exist yet.
Read only data architects already generate (courses, certifications, skills, customer deliveries and milestones) and score every architect on two gates that are never blended: what they've learned and what they've delivered. Experts set the weights; the AI aggregates the evidence; managers validate their own teams and leaders decide.
From guesswork to evidence. The method was reused within weeks for three more cohorts.
Of the model's picks kept after manager review (247 of 254)
Experts staffed into two flagship AI-security delivery offerings
Business units' coverage gaps flagged before staffing, not after
From first request to a leadership-team demo








Every UI visual uses demo data; figures in the copy describe the program itself.
Evidence beats self-report, and guardrails create trust: internal only, never used to judge. The biggest lesson was key-person risk; the model stalled when I moved teams because only I could run it.