Find the value
Choose one consequential workflow where AI can remove friction, improve a decision, or create a capability worth adopting.
01 / What I doIntegration First Workflow Sprint · two weeks · $7,500
The offer
One workflow, made clear enough to build.
Isolate one valuable workflow, define its human and system boundaries, and leave with a decision-ready proof plus a path to production.
Define the boundaries
Map context, actions, permissions, evaluation, handoffs, and the moments that must remain under human control.
Build the proof
Build and test the smallest working system that can earn a real production decision—not a deck pretending the hard parts are solved.
02 / CapabilitiesCross-domain, hands-on, receipts attached
What I bring
Capabilities across the work.
Agentic workflow design
Bounded agents, human approvals, evals, provenance, provider controls, and receipts that survive the demo.
Product & GTM
Product direction, positioning, adoption, and launch systems—from ambiguous opportunity to a decision teams can act on.
Working product proof
React / Next.js, TypeScript, Swift, data pipelines, model APIs, interactive systems, test contracts, and release discipline.
Experience systems
Thirty-five-plus AR experiences, 12 AR-enhanced products, and 3D, motion, Unity, Unreal, and browser-based spatial work.
Operating in the room
Executive translation, cross-functional delivery, enablement, and measured release gates across product, creative, and engineering.
03 / MethodReconstruct → declare → iterate → cultivate
How I work
A technical method, in service of a human outcome.
Reconstruct
Gather the human, technical, cultural, and commercial views before naming the problem.
Declare
Define what should become different before optimizing the work.
Iterate
Build, read the result, compare it with the intention, and revise.
Cultivate
Select the direction that holds and turn it into a durable system.
04 / ProofOperating outcomes · live products · agentic workflows
Commercial proof
Selected evidence.
A business grown 100×. Live AI products. Three agentic operating patterns with the proof level and the unproven edge left visible.
From $10K to $1M in monthly revenue.
I taught myself Blender to bring photorealistic 3D production in-house — cutting roughly $500K a year, funding sharper pricing, and lifting conversion. Campaign content reached tens of millions of views while product, creative, and engineering held to one strategy across US and Beijing teams.
- Role
- ACD / PM — product, creative, and campaign leadership; 12 AR-enhanced products shipped.
- Mechanism
- In-house 3D pipeline → lower costs → sharper pricing → higher conversion, amplified by eight cross-channel campaigns.
- Reach
- Content in the tens of millions of views; single pieces up to 40M.
- Receipt
- Self-reported; full case brief with timeframe and contribution available on request.
LIVE INSTRUMENT / V4.1
VibeScore
Four models. Eight emotional channels. One versioned receipt that exposes disagreement instead of hiding it.
- Angle
- Intent is not impact.
- Artifact
- Multi-model AI instrument + public proof surface.
- Role
- Founder, product strategy, system design, orchestration, interface.
- Limit
- Interpretation, not a verdict or a measure of human worth.
Agentic workflow proof
How the agentic work is bounded.
These are not three chatbots with new names. They are three distinct forms of agency: a live evidence pipeline, an approval-bound publishing agent, and a multimodal evaluation-and-generation workflow.
Context
What the system may know, where it came from, and what must stay attached to the work.
Action
What it may prepare, change, recommend, or execute—and what remains human work.
Permission
Who approves, which destination is allowed, and where the workflow must fail closed.
Evaluation
How quality, disagreement, regressions, and real-world usefulness are measured.
Receipt
The version, inputs, outputs, decision, destination, and result a person can inspect later.
World Wide Vibes evidence pipeline
A scheduled workflow can gather, qualify, score, aggregate, and publish evidence without pretending the automation is the authority.
- Workflow
- Allowlisted sources → bounded ingestion → VibeScore reads → provenance-aware aggregates → public current/history/source views.
- Evidence
- The public product is live, and the repository defines the scheduled pipeline, source floor, scoring basis, release contract, and dated evidence surface.
- Guardrail
- The scorer basis stays labeled as scan, consensus, mixed, or unknown. Editorial approval and production promotion remain separate gates.
- Supports
- Evidence-backed monitoring, research automation, and provenance-aware public data products.
SignalChain approval-bound publishing agent
An agent should be able to prepare the work while the person retains authorship, destination control, final approval, and a usable receipt.
- Workflow
- Supplied idea → six editable formats → revision-bound renders → exact destinations and timing → immutable approval dossier → provider receipt.
- Evidence
- Ninety-four tests pass; shared Apple targets build; the simulator journey verifies all six formats, channel drafts, media bindings, and fail-closed approval controls.
- Guardrail
- No live Postiz connection, physical-iPad journey, external draft, scheduled post, or public URL receipt has been claimed.
- Supports
- Human-in-loop agents, approval systems, permission design, and accountable content operations.
VibeScore CATs multimodal workflow
Model and provider orchestration becomes useful when every transformation can be inspected, withheld, or traced back to its declared input.
- Workflow
- Timed media moments → two emotional arcs → seam evaluation → bridge brief and target arc → optional provider asset → methodology receipt.
- Evidence
- The local Studio and API run end to end, with tests covering the request contract, analysis modes, policy withholding, provider jobs, idempotency, and receipts.
- Guardrail
- Every external generation requires explicit send confirmation. V0.1 is product proof, not evidence that a score predicts audience response.
- Supports
- Multimodal agentic workflows, evaluation design, provider boundaries, and inspectable decision support.
The stack is the practice
A connected working stack.
This is what “strategy that comes to fruition” looks like: one connected practice. Point it at your business — read the market, set the angle, test the work, deploy it, and run the room.
- 01Read the marketWorld Wide Vibes
The emotional and cultural field around your category, made spatial and legible.
- 02Set the strategyAd+Vantage
Audience, feeling, gap, and direction resolved into one portable Strategy Receipt.
- 03Test the creativeVibeScore + the instruments
Compare declared intent with a bounded emotional reading before you spend—and keep the interpretation inspectable.
- 04Deploy the journeyAd+Venture
Turn the campaign into a measured opt-in journey: activation, behavior, and value.
- 05Run the roomAd+Verb Studio
Projects, artifacts, reports, and receipts held in one decision environment.
From the seed of an idea to a decision-ready build your team can evaluate and finish.
Selected surfaces
Current public surfaces.
Three current public surfaces you can inspect without taking my word for it.
World Wide Vibes
A spatial cultural-data product that makes the emotional character of world events visible without reducing them to a flat feed.
View surfaceAd+Vantage
A guided creative-strategy room that turns audience, feeling, gap, and direction into a portable Strategy Receipt.
View surface
Ad+Verb Labs
The company architecture that organizes one emotional-signal engine into strategy, participation, research, and creative workflow products.
View surfaceThe working field
The wider working field.
Every entry names the contribution, the evidence, the boundary, and the paid work it supports. Links open a public surface or an on-page proof record.
Flagship products
The clearest public expressions of the method and its commercial value.
- LIVE PRODUCTVibeScore
A multi-model emotional feedback instrument that returns an eight-channel, versioned reading while keeping disagreement and limits visible.
- Role
- Founder; product strategy, evaluation design, model orchestration, interface, and implementation.
- Evidence
- Live four-model receipt, public proof page, and reproducible validation repository.
- Limit
- Interprets selected content and declared intent; it does not predict audience response or measure human worth.
- Supports
- Creative evaluation systems, AI product strategy, and model-comparison workflows.
- FRONT-END PROTOTYPEAd+Vantage
A guided creative-strategy room that turns audience, feeling, gap, and direction into a portable Strategy Receipt.
- Role
- Product strategy, workflow design, interface, and front-end implementation.
- Evidence
- Openable guided workflow that resolves audience, feeling, gap, and direction into a portable decision record.
- Limit
- Front-end product proof; no claim of completed client adoption or production integration yet.
- Supports
- AI-assisted strategy rooms, decision records, and cross-functional alignment.
- LIVE PRODUCTWorld Wide Vibes
A spatial cultural-data product that makes the emotional character of world events visible without reducing them to a flat feed.
- Role
- Founder; product, evidence architecture, data pipeline, editorial system, interface, and release design.
- Evidence
- Live public product with dated source headlines, measured history, provenance labels, and scheduled ingestion and scoring.
- Limit
- Measures language in a defined news corpus; it does not measure what a population feels or predict events.
- Supports
- Evidence-backed monitoring, provenance-aware AI systems, and public data products.
Engine instruments
Focused tools that apply one emotional-feedback engine to distinct creative workflows.
- LIVE INSTRUMENTImage Analysis
A multimodal workflow that reads campaign imagery, reports its emotional vector, and names the visual drivers behind the result.
- Role
- Product and interface design, multimodal orchestration, and implementation.
- Evidence
- Live instrument returning an emotional vector and named visual drivers for supplied campaign imagery.
- Limit
- A structured interpretation of selected creative, not a verdict or a performance forecast.
- Supports
- Multimodal review tools and creative preflight workflows.
- INSTRUMENT PROTOTYPEStory Composer
A narrative tool that scores individual beats as they are written so a creator can compare the emerging arc with a declared emotional target.
- Role
- Product concept, narrative workflow, scoring integration, and interface.
- Evidence
- Openable beat-level editor comparing an emerging emotional arc with a declared target.
- Limit
- Instrument prototype; it does not predict story quality, preference, or commercial performance.
- Supports
- Human-in-loop creative tools and real-time decision support.
- WORKING PRODUCT PROOFVibeScore CATs
A content-aware transition workflow that measures two timed media arcs, evaluates their seam, and produces an inspectable bridge brief and receipt.
- Role
- Product strategy, versioned contract, gateway, provider boundaries, browser studio, and validation design.
- Evidence
- Working local Studio and test suite cover timed-media analysis, seam evaluation, bridge generation, provider consent, and the final receipt.
- Limit
- V0.1 under validation; there is no claim that transition scores predict human preference or campaign outcomes.
- Supports
- Multimodal agentic workflows, evaluations, and consent-bound provider integrations.
Systems and ventures
Products that connect strategy, participation, artifacts, and decisions into larger systems.
- PROTOTYPE WORKSPACEAd+Verb Studio
A workspace that connects projects, artifacts, emotional reads, Bridge Reports, and receipts into one decision-making environment.
- Role
- Product architecture, workflow design, information model, interface, and implementation.
- Evidence
- Openable workspace proof connecting project context, artifacts, reports, emotional reads, and decision receipts.
- Limit
- Prototype workspace; client-team adoption and production operations are not yet proven.
- Supports
- Decision environments, context architecture, and AI workflow adoption.
- ACTIVE PILOTAd+Venture
A participation-planning system that turns campaigns into opt-in journeys with observable activation, behavior, completion, and value exchange.
- Role
- Venture concept, participation model, product strategy, interface, and implementation.
- Evidence
- Openable pilot surface mapping attention, opt-in, action, completion, and value exchange as separate stages.
- Limit
- Active pilot; the planning system does not yet prove campaign lift or participant demand.
- Supports
- Participation journeys, activation design, and measurement planning.
- VERIFIED WORKING PROOFSignalChain
An approval-bound Apple workflow that turns one supplied idea into six editable campaign formats and an exact-destination delivery dossier.
- Role
- Product strategy, native architecture, approval contract, implementation, and QA.
- Evidence
- Ninety-four passing tests, shared Apple target builds, and a verified simulator journey with revision-bound media and exact-destination controls.
- Limit
- No live Postiz connection, physical-iPad exercise, or external publishing receipt yet.
- Supports
- Human-in-loop agentic deployment, permissions, approvals, and audit trails.
- ACTIVE COMPANYAd+Verb Labs
The company architecture that organizes one emotional-signal engine into strategy, participation, research, and creative workflow products.
- Role
- Founder; portfolio strategy, shared evidence architecture, product systems, brand, and GTM.
- Evidence
- Public company surface connecting live products, openable prototypes, research, and one shared receipt model.
- Limit
- A coherent product ecosystem and operating practice; breadth alone is not evidence of market demand.
- Supports
- Multi-product architecture, venture studios, and product-to-market systems.
Research and proof
Public methodology, reproducible tests, and evidence supporting the product thesis.
05 / ContactDiagnostic · sprint · team
Contact
Start with the problem.
Workflow diagnostic.
One workflow truth map: where context breaks, which actions should be automated, what must stay human, and whether a build is justified.
Scope the diagnosticIntegration First Sprint.
One bounded workflow, a working decision-ready proof, evaluation plan, permission model, and a production handoff your team can use.
Scope a sprintPut me in the room.
AI workflow deployment, product and GTM strategy, creative technology direction, and high-leverage special projects.
Discuss the role