Internal Go-Forward Strategy

M Hudson StudioGo-Forward Rollout Plan

Version v1 Baseline · Pre-Kickoff From Jonathan · Fractional AI Architect/Director (aiChemist) Audience Internal — Matt + Carlos + Jonathan Date 2026-06-19
v1 Baseline — not the verdict

The kickoff call has not happened yet. Every role boundary and sequencing claim below is a proposed hypothesis to validate live on the call — not a settled decision. The companion alignment form is the instrument that resolves each open question.

Carlos builds the worlds; Jonathan builds the machine that turns those worlds into a repeatable, branded, never-fails client experience. We meet at one clean seam — Carlos stages output samples, Jonathan ingests them into the pipeline.
01 · The Frame

One base plan, three people aligned

This is the shared go-forward map for the first stretch of the engagement: how the work sequences, who owns which lane, and how it tracks against the studio's milestones. It is a starting position, drawn to be confirmed or adjusted in the room.

Jonathan is across the pipeline (Lane C — AI orchestration and the "one pane of glass" portal). Carlos is inside the 3D craft (Lane B — TwinMotion/Unreal execution). The only place the two lanes touch is the sample handoff. That single seam is the whole boundary — proposed here, confirmed on the call.
02 · Proposed Role Split

The bright line (a hypothesis)

Drawn so the line is visible up front rather than negotiated mid-stream. Treat the table as a proposal — the alignment form lets you confirm it as-is or move the seam.

Lane B · In-Room Operator
Carlos
  • TwinMotion / Unreal execution
  • Gaussian-splat model training
  • Offshore modelers + quality control
  • Lives and breathes the 3D craft
Lane C · Fractional AI Architect
Jonathan
  • Inputs → ops → outputs pipeline
  • Branded "one pane of glass" portal (on Matt's server)
  • AI concept-generation workflow (ComfyUI)
  • The layer that stays aware end to end
The seam — one handoff: Carlos stages output samples → Jonathan ingests them to prep the pipeline and the presentation. No overlap by construction.
Proposed concept-workflow pipeline — where the two lanes meet
03 · Live Momentum

What is already running

Functional today — named honestly, including the piece still in progress.

Functional

Chain A — Variant Studio

Drawing → render with structure-lock; commercially-licensed model. ~33 renders. Jonathan's track.

Functional

Chain D — Landscape Inpaint

Masked inpaint, house/pool pixel-faithful. ~12 renders. This is the "3–4 design themes per property."

Functional

Chain B — Splat Flythrough

Real splats running in-browser. A running start for Carlos's track, not a flag-plant.

In Progress

Chain C — 360 Vibewalk

Viewer shell present, assets not yet populated. Named as the next piece — not oversold.

Live momentum — chain status at a glance
04 · Sequencing

The order of operations

Proposed priority order, parallel-tracked so the plan survives travel and hardware timing.

1. Gaussian splatting → 2. AI / TwinMotion concepts → 3. Unreal Engine (evolves naturally after the first two boxes are checked). Carlos heads-down on splat training while Jonathan + Matt stand up the AI concept workflow on live and legacy jobs in parallel — complementary, and the concept interface is reusable into the advanced portal.
05 · Timeline

Built around Matt's travel

A parallel-track plan that holds through the July windows toward the studio milestones.

Baseline timeline — validate dates on the call
Milestones: compelling demo mid-August · real deal — first live session mid-October · grand opening November 1. Site inventory in Jupiter tentatively Thu 6/25 (two projects; 360 + drone if lidar slips). Matt's July: San Diego ~1wk → back 2wks → China ~2wks.
06 · Assumptions

To validate on the call

Everything above rests on these. Each maps to a decision on the alignment form — they are hypotheses, not conclusions.

  1. Output staging — samples stage on Matt's in-house server.
  2. Beta set — the two Jupiter 6/25 projects are the beta case studies.
  3. Sequencing — splatting → AI/TwinMotion → Unreal holds.
  4. Cadence — a 3–6 month value-gated runway is acceptable, dialed both ways.
  5. Contract — first payment complete 6/19.
  6. Beta data — two project folders + the phase/stage process-flow arrive via shared drive to unblock concept work.
How the baseline resolves into v2
07 · Confirm the Plan

Six decisions, one form

The alignment form converts the six open questions into a quick confirm-or-adjust pass. The recommended option is pre-highlighted on each — push back on any of it.

Confirm the plan → Alignment form