Dev Plan v0.3
Fluxx Development Plan v0.3 is the master document that defines what gets built, in what order, by whom, with what money, on what timeline. v0.3 supersedes v0.2 and earlier dev-roadmap budgets. Source:
02_Working/Strategy/Dev_Roadmap/Fluxx_Development_Plan_v0.3.docx.
The dev plan organizes the Fluxx build into five categories that run in parallel where bandwidth and dependencies allow. Categories are internal organizing language; they do not surface on investor-facing decks. The cost envelope is $6.4M to $12M build plus $2.4M to $4M per year maintenance.
The five-category framework
The framework provides the structural decomposition; each category owns a coherent module set with its own scope, vendor potential, and sequencing. Full per-category specs at Build Categories.
| Cat | Name | Build profile | Y1-Y5 outsourced |
|---|---|---|---|
| 1 | DCC Pipeline Framework | Core architecture plus per-DCC outsourced | $200K / $700K / $700K / $100K / $0 |
| 2 | Fabrixx Virtual Cloud Infrastructure | Core architect plus DevOps outsourced plus TPN audit | $0 / $500K / $500K / $100K / $0 |
| 3 | Fluxx Automate | Core team plus GlassBox (Norman V7) | $406K V7 remaining (Y1), then $0 / $300K / $400K / $100K / $0 |
| 4 | Agentic Build Assist (HEAVY INVEST) | Core 2-3 dedicated TDs plus per-DCC specialists | $100K / $500K / $800K / $400K / $200K |
| 5 | AI Token Budget (Opex) | Recurring opex, scales with team size | $131K / $306K / $438K / $525K / $613K |
Cat 4 is the highest sustained outsourced line; the plugin SDK and agentic harness quality determine the third-party adoption ceiling. Cat 5 is opex (not build), scales linearly with developer headcount, and is structural to how the engineering team operates.
Team architecture
Core in-house equity-bearing team plus outsourced vendor model (per-module RFPs, scoped engagements). Y5 placeholder reaches 15 named roles at full scale (8 to 10 of those are engineering-architecture core; the rest are founders, creative supervisors, on-set ops, sales). Full per-role specs at Team Architecture.
Architectural locks for the team layer:
- Ari Jampolsky title: Executive Director & Co-Founder. S126 reverted prior S123 “Director of Set Operations” lock; cascade complete across canon surfaces.
- No third-party at the core of the stack. Production tracking built in-house under Fluxx Automate (Cat 3), not via ftrack or ShotGrid integration at the runtime layer.
- Norman / GlassBox is vendor only. Not partner, not co-founder, not deferred-comp. The v1.4 JV-shaped financial model had a “CTO (Norman)” line with cash plus deferred comp; v2.0 (Fluxx-standalone, Q-G89) removes that line and treats Norman as Cat 3 vendor spend.
Y2 hires at Series A close (Jan 2027): CTO, Pipeline architect, Cloud architect, Automate product owner, two Senior Pipeline TDs, Integration engineer. Mentor / advisor layer: Jeff Brown (Q-G9), Bill Polson (Q-G10), Brian Connor.
Build sequencing (Y1 to Y5)
Y1 (2026)
Existing seed runway. Cat 3 Norman V7 remaining work ($406,070 spread across Y1) plus first Cat 1 modules ($200K) plus first Cat 4 specialist contractors ($100K) plus proving-ground demo ($450K). Cat 2 and Cat 3 forward modules do not start (gated on Series A). Cat 5 AI tokens $131K (3 devs midpoint).
Y1 cash gap. The $500K starting cash placeholder leaves a $1.9M shortfall by end of Y1 against the spend profile above. Three resolution paths surfaced in the financial model dashboard: (1) confirm actual Jan 2026 cash balance (placeholder may understate), (2) reduce Y1 outsourced spend (defer Cat 1 starts, scale back the proving-ground demo, push Norman M4 plus M5 to Y2), (3) bridge raise or convertible note in 2026. A fourth path is implicit: accelerated Series A close earlier than Jan 2027. See v2.0 Financial Model for the math.
Y2 (2027)
Series A close Jan 2027 ($10M placeholder; industry median 2026 is $15M). CTO joins. Pipeline architect, cloud architect, Automate product owner, two senior pipeline TDs, integration engineer all start. Core team scales to about 10 FTE. Outsourced modules ramp: Cat 1 $700K, Cat 2 $500K, Cat 3 forward $300K, Cat 4 $500K. First revenue $2.4M. Vendor RFPs out for Cat 1 (Nuke plus Unreal plus Maya first), Cat 2 (Cloud DevOps build-out, TPN-aware), Cat 4 (per-DCC specialist contractors).
Y3 (2028)
Full team scale (8 to 10 FTE engineering-architecture core). Third senior pipeline TD added. Sales / BD lead added. TPN audit one-time ($150K), gates Sony / A24 / Netflix production tier. Outsourced spend peaks: Cat 1 $700K, Cat 2 $500K, Cat 3 $400K, Cat 4 $800K. Revenue $6.5M. EBITDA inflects positive. Maintenance phase factor kicks in (25 percent of cumulative Y1-Y2 outsourced spend, applied flat Y3-Y5, $677K each year).
Y4 (2029)
Productions at scale. Cat 1 outsourced spend drops to $100K (long-tail). Cat 4 contractor pool $400K. Revenue $16.5M. Hardware capex $2M (outright stage gear, transitioning from loan/lease).
Y5 (2030)
Mature run rate. Most categories in maintenance plus iteration. Cat 4 contractor pool $200K (maintenance retainer level). Cat 5 AI tokens $613K. TPN re-audit $80K. Hardware capex $3.5M. Revenue $31.1M. 24 productions served. 12 hardware packages active. EBITDA $21.4M, gross margin about 91 percent (Y5 dashboard snapshot).
Maintenance phase
Maintenance begins as each category’s initial build ships. By Y3 the whole stack is in maintenance plus iteration mode. Annual cost shape is roughly 25 to 30 percent of initial build cost annually (typical for SaaS-style infrastructure products), totalling $2.4M to $4M per year plus cloud opex.
Maintenance covers:
- Bug fixes plus stability. Per-module bug fixing across DCC plugins, cloud infra, Automate backend. Incident response. Performance optimization.
- DCC version compatibility (load-bearing). DCC tools release annually (Maya 2027, Nuke 16, Houdini 21, Unreal 5.x, Blender 4.x). Each major release requires plugin compatibility work, days to weeks per plugin. Long-tail support for productions in flight on older DCC versions.
- Security plus compliance. Ongoing TPN re-audits (annual). Security patches as upstream vulnerabilities surface. Compliance with evolving production standards (Sony, A24, Netflix policies, EU AI Act).
- Customer plus production support. Production-specific integration help. Documentation maintenance. Office hours, escalation path, on-call rotation.
- Feature evolution. Roadmap items deferred from initial build. Customer-driven feature requests. AI / ML model updates.
Team architecture shifts in maintenance: core team slims slightly. Initial-build architects move to product leadership. Some become full-time maintenance and customer engineers. Outsourced module relationships shift from project-based to retainer-based (continuity matters; the original module builder is the most efficient retainer).
Maintenance run rate breakdown:
| Line | Annual |
|---|---|
| Core team (slimmed) | $1.4M-$2.0M (7-9 FTE) |
| Outsourced module retainers | $600K-$1.2M (20-30 percent of build cost) |
| AI token budget (Cat 5) | $300K-$600K |
| Cloud / GPU provider opex | TBD (scales with production load) |
| TPN re-audit (annual) | $40K-$80K |
| DCC license plus SDK fees | $30K-$80K |
| Customer support tooling | $30K-$60K |
| Total maintenance run rate | $2.4M-$4M plus cloud opex |
Once productions adopt Fluxx, the run rate is covered by IaaS plus per-shot revenue. The model is for maintenance to be self-funding from Y3 forward.
Norman V7 M1-M5 as Cat 3 Automate timeline
Norman / GlassBox is the Cat 3 Automate vendor for the V7 milestone series. Stated cumulative V7 total: $517,000. Paid through V7 to date: $85,730 (M2 plus M3 partial). V7 not-yet-invoiced: $406,070. Source: 03_Fluxx_Private/Legal/Norman_V7_Milestones_Review_2026-05-12.md.
| M | Name | Window | Status | Stated | Cumulative |
|---|---|---|---|---|---|
| Proto | Prototyping (Fabrixx plus Vault) | Pre-Demo 1 (2024-2025) | Completed | $20,000 | $20,000 |
| M1 | Fabrixx MVP for Technical Test | Nov 2025 | Completed | $119,830 | $139,830 |
| M2 | Fabrixx Improvements (Automate framework initial) | Nov-Dec 2025 | Completed (PAID) | $56,930 | $196,760 |
| M3 | ComposiTrack Polish plus Automate prototype | Dec 2025 - Jun 2026 | Partial | $50,400 | $247,160 |
| M3.5 | Test Shoot 2 (Unreal switch, Automate hardening) | May-Jun 2026 | In progress | $16,050 | $263,210 |
| M4 | Fluxx Automate Cloud (FE/BE/DB split, job queue, AI QC) | Jun-Nov 2026 | Planned | $110,870 | $374,080 |
| M5 | Fluxx Tool Suite Production Readiness (hybrid compute, agentic telemetry, queue mgmt, preliminary TPN) | Nov 2026 - Mar 2027 | Planned | $142,920 | $517,000 |
M3.5 engine direction shift. Chaos V-Ray Arena → Unreal Engine. Driven by difficulties at the December test shoot. Affects pitch story (Glossary previously listed Chaos V-Ray Arena as the renderer), ChromaTrack patent implementation (V-Ray Arena GPU-framebuffer architecture diverges from Unreal nDisplay post-process shader), and JV positioning. Kevin’s deliberate call. Time-sensitive.
M4 scope expansion (about 6x). Prior framing was $18,200 plus UI TBD; V7’s new M4 is the cloud platform build at $110,870. Kevin must explicitly authorize against the V7 framing. M4 Item 1 (Initial Cloud Readiness, $28,950, 300 hrs) is the super dependency: splits Fluxx Automate from local desktop to cloud-native FE/BE/DB. Every later milestone (M4 Items 2-4, all of M5) depends on this. Kevin’s acceleration ask: Option B (about 100-130 hrs in May 2026).
M5 TPN gap. Preliminary TPN only; full TPN audit explicitly out of scope and needs separate budget line ($80K-$200K per dev plan).
M5 cloud provider hardcode. AWS EC2 is hardcoded in M5; per Q-G49 Fabrixx is the canonical cloud-side compute layer, with EC2 as one implementation option among multi-cloud abstractions.
Open items (Kevin gap-fill)
Cross-walk of dev plan §13 plus financial model Notes tab plus dashboard explicit flags.
Time-sensitive
- Engine direction call (Unreal vs Chaos V-Ray). M3.5 in flight.
- M4 scope authorization to Norman. V7’s $110,870 vs prior $18,200 framing.
- M4 Item 1 acceleration ask. Option B for May 2026.
- M3 commercial state reconciliation. Patents line $3,600 status. V7 numbering items 2 and 4. Paid / outstanding / deferred since April 22 snapshot.
- ChromaTrack vs ComposiTrack patent scope. Counsel needs answer before PCT decision Mar 31, 2027. ComposiTrack canonical per Q-G88.
- JV go / no-go decision. Affects whether to fork a JV-specific financial model variant from v2.0.
Team architecture
- Core team hiring sequence (Pipeline architect first vs Cloud architect first).
- TD hiring profile (generalist senior vs specialist per DCC).
- Outsourcing geography (local studios, US-only, global).
- Vendor RFP process (formal small RFP per module, or relationship-based).
- Norman / GlassBox retention beyond V7.
- CTO 2027 search. Founder-market comp benchmark needed.
- Mentor section to add to Headcount tab in v2.0.
Architectural (gate downstream work)
- Priority DCC list for v0.1 (Nuke + Unreal + Maya working assumption).
- Plugin SDK scope at v0.1 (public for third-party studios or internal-only).
- USD posture (native vs aware).
- Multi-tenant VPC topology (per-tenant vs shared with tagging).
- Cloud / GPU provider stack pick (Q-G49 §7 parked).
- Render farm manager at v0.1.
- Remote desktop protocol (Teradici, Parsec, NICE DCV, native).
- Pre-comp recipe serialization (Nuke fragment, USD layer, custom JSON).
- Per-shot cost attribution algorithm.
- Velvet to Fluxx pipeline cover relationship.
Financial model
- Y1 starting cash ($500K placeholder vs actual Jan 2026 balance).
- Series A ask ($10M placeholder vs industry median 2026 $15M).
- CTO comp band (founder-market data needed).
- Outsourced module $ (rough allocations vs vendor RFPs).
- Cloud / GPU opex (scaling placeholder vs provider pick).
- AI token mix ($175/dev/day midpoint vs first-few-months actual usage).
- Equipment loan/lease ($500K-$2.5M cumulative placeholder).
Source dependencies
02_Working/Strategy/Dev_Roadmap/Fluxx_Development_Plan_v0.3.docx(primary)02_Working/Strategy/Dev_Roadmap/Fluxx_Fusion_Dev_Roadmap_Budget_v1.xlsx(predecessor)02_Working/Strategy/Dev_Roadmap/Fluxx_Fusion_Dev_Roadmap_Budget_v1_1.xlsx(predecessor refinement)02_Working/Financial_Models/v2.0_Fluxx_Financial_Model.xlsx(cost envelope)03_Fluxx_Private/Legal/Norman_V7_Milestones_Review_2026-05-12.md(Cat 3 vendor reference)
Audience visibility
investor, counsel.