When the information is fragmented and nobody owns the whole picture, I externalize it, name what connects to what, impose a structure, and hand that structure to people who keep operating it after I stop touching it.
Six documented eras, 2010 to 2026, across four industries and every change of title and technology. In 2014 an employer put it in writing, company-wide, twelve years before it became a job title.
01 / WHAT I ACTUALLY DO
My background is not in coding. What I have instead is a sixteen-year artifact trail, and an employer who named the behavior in writing before it had a job title.
I land in a commercial environment nobody has mapped, convert it into an explicit artifact — a playbook, a pricing rule, a framework, a capture system — and assign it to named owners. The test I hold myself to is whether it still runs after I stop touching it.
Promoted “not only for successfully closing significant revenue in new Digital and Field business but also by infusing business development mentality into many pockets of the organization through coaching, training and mentorship.” Note the conjunction: the transfer behavior earned a promotion when it was paired with revenue. On the record as it stands, it has never earned one alone.
Where the playbook already exists, is owned by the firm, and the job is to run it at volume, there is nothing to structure — and I look like everybody else in that seat. That is a real boundary, it is documented below, and it is the most useful thing on this page for deciding whether to call me.
AI deployment strategist. Forward-deployed strategist or operator. Applied AI transformation lead. AI operations or deployment lead. The reporting line is not the point; what matters is organizational ownership, cross-functional influence, real AI adoption responsibility, and ambiguous, consequential problems. Not a quota. Not a software-engineering seat.
02 / WORK & EVIDENCE
Evidence beats adjectives. Nothing here is upgraded by being written down.
Problem. Lost the Sirius XM account. The stated reason — “too expensive” — was diffuse and unusable as given.
Artifact. An unrequested post-loss memo to my manager, 2015-05-22, that turned the loss into a pricing rule: they were ~20% more expensive across the board but up to 80% on a few specific line items, so the fix was to charge more for what buyers already expect to pay a premium for, reduce investment levels on what they are less familiar with, and restructure the Playbook around how the market actually budgets.
Who used it, and what became reusable. The firm. Fourteen months later, 2016-07-21: “Slide 30 will include our 2 alternate CCS recommended models and will set us up for the pricing discussion.” A pricing posture outlived the account that produced it.
Does not prove: that I owned pricing strategy, or that the change produced measurable revenue. The memo and the later deck both exist; the causal link between them is an inference I am not making.
Problem. A sales function with no defined process and no system of record.
Artifact. “Q3 Sales Goals Framework, Screen shots and Project Plan” — named owners, dated deliverables, designated systems of record, and a self-auditing loop: the team spends five minutes per person in each weekly sales meeting on their pipeline, and I compare that against the meeting report to confirm every meeting was actually recorded. Companion artifacts: a documented sales process, a weekly game-plan template, an ROI tool instantiated for four clients.
What became reusable. The process and its templates, held by named people rather than by me.
Does not prove: sustained adoption after my ten months there, which I did not observe and do not claim.
Problem. A newly legal, fragmented dispensary channel — raw regulator data, and no route from it to owned work.
Artifact. “Dispensary Playbook 2021” — 97 steps, each carrying a DELIVERABLE tag against a named colleague: “Atosa and Luz to drive on creating this, Eric to provide feedback.” Operating frame: “We sell, we train, we promote and we stay connected.”
What became reusable. A channel-entry procedure other people owned by name, including steps I explicitly did not own.
Does not prove: revenue outcomes, or that every step was executed.
Sold and helped scope enterprise professional-services engagements, partnering with Adobe consulting and technical specialists through client-needs discovery, transformational workshops, and engagement stand-up. Named on Adobe's internal Customer Spotlight for the HSF Affiliates win — a three-year AEM Managed Services engagement with Berkshire Hathaway HomeServices — alongside solutions consulting, technical solution architecture and partner leads. Co-facilitated 10 Customer Solution Design Workshops.
Does not prove: delivering the implementations myself, or any later AI adoption outcome. No revenue figure is published here: the commission record exists and has not been opened, and until it is, the number stays off this page.
Held cross-functional responsibility across sales, ecommerce, marketing, customer experience, and operations, partnering with the CEO on strategy and AI-assisted operating methods. Mapped prospecting, qualification, outreach, and follow-up across CRM, ecommerce, support, and compliance, and designed AI-assisted routines with human review before consequential external actions.
Does not prove: measured AI return on investment, or adoption beyond one company.
This is the most recent instance of a sixteen-year-old habit, not a new direction. What is actually new is scale and self-instrumentation: an append-only decision ledger that freezes decision-time confidence, resolves outcomes against an external source rather than memory, and derives how much to trust an agent from how often I override it. The behavior is 2010 vintage. The tooling is three months old.
It spans Claude, Claude Code, Codex, and GitHub and runs my own strategic and operational work: research, planning, artifact production, and action-state tracking. Built around reusable instructions, explicit success criteria, decision logs, and designed review points. Its checking layer has caught stale sources, duplicate work, and unsupported language before I acted. Day to day that is judgment orchestration: choosing the model and the tool, sequencing the work, setting the evidence and authority gates, and seeing a commitment through to a verified finish line.
Does not prove: team or customer adoption. It is a personal system, and I don't claim independent software-engineering skill.
Managed executive search delivery across cannabis and beverage clients: role scoping, sourcing, interview-process coordination, and offer support. Hosted the Cannabis Matters executive interview series.
ForceBrands had substantial operating structure before I arrived — onboarding and training materials, a sales playbook, the 8 Steps of Recruiting, an intake-call deck and cheat sheet, job-lead stages, named owners — documented in a curriculum created three weeks before my start date. Executive search delivery is a pre-structured job: the playbook exists, the firm owns it, and the work is to run it at volume. In four years there I built far less that was reusable than in ten months at Dark Heart.
A pre-structured environment is a lower structure-creation opportunity. That is the boundary condition on everything above, and it is why the roles I want are the ones where building the structure is the job rather than something that happens beside it.
Does not prove: that I cannot operate in a structured environment. It proves the mechanism has a precondition.
Two operating principles entered public testing and were killed on 2026-07-31, causes of death on the page. One claim is under test with its thresholds locked before any data. You can read every entry, dated, in the ledger.
Does not prove: anything about anyone's business but my own.
03 / THE RECORD — A PUBLIC WORKSHOP FOR COGNITIVE HUMAN INFRASTRUCTURE
Every claim gets its date before the outcome exists. Every test gets its rules locked before any data shows up. When an idea dies, the kill stays on the page with a cause of death.
RATIFIED 0 UNDER TEST 1 KILLED 2
So far? Two ideas went into testing. Both died. Zero survivors. Zero survivors doesn't mean the work failed. It means the bar held.
UNDER TEST The record can carry the name 2026-08-02
KILLED Finishing > Starting Cause of death: a real principle does not need forty reminders. 2026-07-31
KILLED Receipt > Report Cause of death: pointed at other people, it reads as surveillance. 2026-07-31
Notice what keeps recurring. Pressure-test it. Publish only what refuses to die.
State the claim before the result. Lock the window. Let the outcome win.
Keep the misses on the page. Especially the misses. A clean record is usually a dirty trick.
04 / THINK
Thought leadership without the thought costume. Claims get dates. Predictions get windows. Wrong ideas get excellent obituaries.
Knowledge accumulates. Wisdom compresses. Behavior compounds.
Read itArchitecture has a dangerous property. It always feels productive. You can spend weeks designing the perfect system without moving the original mission forward an inch.
Read itEvery repetitive manual task is a map. It’s showing you where the next piece of infrastructure doesn’t exist…yet. Every frustration is a clue, showing you a breadcrumb for...
Read it05 / MUGS / OBJECTS UNDER TEST
One mug. One color. A changing field of prompts built to make the next useful move obvious before your coffee gets cold.
Reality is the only stakeholder with permanent veto power. This white 11 oz Flowkeeper Mug asks the useful question before another five minutes disappear:...
Thinking without context is expensive guessing. This white 11 oz Flowkeeper Mug reminds you to load what is open, blocked, already decided, and first...
Your ego has excellent publicists. Evidence does not need one. This white 11 oz Flowkeeper Mug puts the primitive where the meeting, model, or...
THE SELECTION LOOP
No manufactured hero SKU.
What people use gets another round.
Demand earns permanence.
Only after they are gone do they become past limited editions.
No countdown clocks. No scarcity theater. No saving a slogan because I have feelings about it.
THE SMALL PRINT, MADE LARGE
A former cannabis breeder who learned not to argue with the harvest, now applying the same pressure to ideas, agents, and work.
Dates, test windows, misses, revisions, and kills stay visible. The method is part of the product.
Reasoning and resolved positions can become public. Family, confidences, clients, and open negotiations do not.
WHAT TO DO NEXT
And if you read something here and think I'm wrong, good. Tell me.