AC
Aigarth Tracker
Phase 0 — Evaluation
Use case study

Video Synthesis ANNs

A coordinator-based approach to AI video: a small Director ANN plans the shot list, then Camera, Motion, Depth, FX, Audio, and Quality ANNs do their part. Each is a published, versioned, rateable module in the Aigarth marketplace. This page is the live view of the proposal, the architecture compatibility verdicts, the development roadmap, the cost model, and the risk register.

Specialized roles
7

Director, Camera, Motion, Depth, FX, Audio, Quality

New compute kinds
6

render, motion, depth, fx, audio, quality

New tables
5

ann_pipelines, ann_pipeline_runs, ann_feedback_events, compute_workers, ann_version_metrics

Phase 0 (now)
~75 min

evaluation, blog, dashboard, tracker

Architecture compatibility
How well the current Aigarth Cloud architecture can support each pillar of the proposal, and what we need to add.
ANN compatibility
services/ann
Extend8/10

ANN registry, versioning, reviews, marketplace listings, deploy hook — all ready in services/ann. We need a `role` enum, a `pipeline` table, and a `pipeline_runs` table to model Director/Camera/Motion as first-class roles.

Gap

Add `ann_role` enum, `ann_pipelines`, `ann_pipeline_runs`, `ann_feedback_events`. ~3 new tables, 6 new endpoints. No existing flow changes.

Compute architecture
services/compute
Extend7/10

services/compute already does typed jobs, clusters, regions, reservations, and idempotency. The worker-side story is greenfield — we need a worker protocol, capability advertisement, and sandboxed execution.

Gap

Add `compute_workers` table, extend `jobs.kind` enum with 6 new kinds (render, motion, depth, fx, audio, quality), add `/v1/workers/*` namespace + long-poll `/v1/jobs/next`.

Continuous learning
services/ann + services/compute
Greenfield5/10

Nothing in the current system stores per-ANN quality signals, user feedback attribution, or version-over-version improvement. This is the largest single gap and the part that turns the use case from a demo into an evolutionary system.

Gap

Add `ann_feedback_events` table (kind, value, attributed to ann_version + pipeline_run) and a `ann_version_metrics` materialized view. Cron job to refresh. ~1 new table per service, 1 rollup job.

Qubic integration
services/qubic
Blocked9/10

The shape is right — services/qubic already has wallets, stakes, treasury, validators, network. On-chain ANN ownership, worker registry, and reputation map cleanly to Qubic smart contracts. But K12 signature verification is format-only today, so on-chain writes are not yet safe.

Gap

Ship real K12 signature verification (precondition). Then deploy on-chain `ann_registry`, `worker_registry`, `reputation`, `rewards` contracts. Off-chain rollup commits to chain once per day.

Specialized ANN team
Seven small ANNs collaborate to produce a video. Each is published, versioned, and rateable in the Aigarth marketplace.
Designed
Director

Reads the prompt, plans the shot list. Returns a structured JSON describing scenes, durations, camera, audio, and style.

(LLM call, not a compute job)
Designed
Camera

Turns each shot into a 3D camera path through space, with lens parameters. Knows 50 cinematic moves and when to use each.

(rule-based, no GPU)
Designed
Motion

Generates in-between frames from a keyframe and a camera path. Pure interpolation, deterministic, fast on CPU.

motion.frame_interpolate
Designed
Depth

Produces per-frame depth maps. Foreground / background separation for parallax and depth-of-field. Uses MiDaS on CPU.

depth.midas_small
Designed
FX

Composites particles, steam, light glows, atmosphere. Domain-specific — a Steam FX ANN and a Rain FX ANN can coexist.

fx.composite
Designed
Audio

Aligns narration, music, and sound effects to the timeline. Knows when to dip the music under a voice.

audio.align
Designed
Quality

Watches the result. Spots artifacts, weird motion, dropped frames. Decides whether to ship or to flag a stage for re-run.

quality.evaluate
Render flow
promptDirectorCameraMotionDepthFXAudioQualityRenderMP4

Every stage emits a structured event into ann_feedback_events. The Quality ANN at the end can flag any stage for re-run; only that stage re-executes, not the whole pipeline.

Roadmap
Five phases. Re-evaluate at the end of Phase 1 with a real 30-second render before committing to Phase 2.
Phase 0DocumentationIn progress

Architecture evaluation, blog article, this dashboard. No code in production.

  • Architecture compatibility evaluation (4 dimensions, color-coded)
  • Blog article on the Aigarth Cloud website
  • Dashboard /video page live with the architecture diagram
  • Risk register + cost model
Exit criteria

Stakeholders aligned. Decision: proceed to Phase 1.

Phase 1Centralized prototypeNext

Single Python process acting as Director + Camera + Motion + Render. A prompt in, an MP4 out, every stage recorded.

  • `role` enum + `ann_pipelines` + `ann_pipeline_runs` tables
  • Director ANN (LLM call)
  • Camera ANN (rule-based)
  • Motion ANN (OpenCV affine transforms)
  • Render ANN (FFmpeg)
  • Mock pipeline-run endpoint
  • 30-second sample render: 'city at sunset'
Exit criteria

30s MP4 from a 1-paragraph prompt in under 10 minutes. Decision gate.

Phase 2ANN marketplace integration

Publish ANNs as real listings. Sell pre-composed pipelines. Wire up per-version metrics and reviews.

  • Director / Camera / Motion / Render ANNs published as listings
  • `kind = 'pipeline'` listing type
  • `ann_version_metrics` materialized view + cron
  • Per-ANN quality ranking in the marketplace
Exit criteria

Users buy and run a 5-ANN pipeline end to end without help.

Phase 3Distributed compute

Worker registry, worker protocol, `aigarth/worker-video` image. Real per-stage cost.

  • `compute_workers` table + `/v1/workers/*` namespace
  • `POST /v1/jobs/next` long-poll + `POST /v1/jobs/:id/progress`
  • `aigarth/worker-video` Docker image (Python + OpenCV + FFmpeg + PyTorch CPU)
  • Per-worker reputation scoring from feedback events
  • 3-worker local cluster for E2E testing
Exit criteria

Workers earn reputation; misbehavior is slashed; per-stage cost is real money.

Phase 4Qubic ecosystem

On-chain ANN ownership, worker registry, reputation, rewards. Off-chain rollups committed to chain.

  • Real K12 signature verification (precondition)
  • On-chain `ann_registry`, `worker_registry`, `reputation`, `rewards` contracts
  • Daily off-chain rollup → on-chain state root commitment
  • QUBIC reward distribution: 70% worker, 20% ANN author, 10% treasury
Exit criteria

An ANN's ownership and a worker's reputation are verifiable on-chain.

Cost model (illustrative)
30-second 1080p render, all stages run once, single CPU worker. Numbers are placeholders that will change as we ship and measure.
StageCost (QU)LatencyNote
Director (LLM call)0.052 sOne round-trip to gateway
Camera (rule-based)0.01< 1 sPure function
Motion (OpenCV, CPU)0.5060 s900 frames × 65 ms/frame
Depth (MiDaS, CPU)0.3045 s900 frames × 50 ms/frame
FX (compositing)0.1010 s900 frames × 11 ms/frame
Audio (alignment)0.055 sOne short clip
Quality (model)0.1015 sOne full pass
Render (FFmpeg)0.0530 sEncode h264
Total~1.16~3 minper-stage cost visible to user before commit

The cost is dominated by Motion and Depth (every frame). Per-stage pre-flight estimates let the user decide: pay for a faster worker, swap in a cheaper ANN, or skip the stage.

Risk register
Top 8 risks, ordered by impact × likelihood. L = likelihood, I = impact.
#1
Video is expensive — 5–10 min on CPU, 30–60s on a single GPU

Mitigation: Per-stage pre-flight estimate, per-stage credit deduction, pre-purchased capacity via reservations.

L:highI:high
#2
Multi-ANN coordination compounds errors — bad Director cascades

Mitigation: Quality ANN at the end of every pipeline; per-stage retry; A/B two Director ANNs in shadow.

L:highI:high
#3
Worker trust — untrusted code on shared infrastructure

Mitigation: Docker sandbox; no network egress except Aigarth APIs; resource caps; reputation + slashing.

L:medI:crit
#4
Feedback attribution — when a pipeline fails, which ANN is responsible?

Mitigation: Per-stage `attribution` field on each feedback event; Quality ANN emits a 'weakest stage' verdict.

L:highI:med
#5
Latency — 30s render in 5 min is bad UX

Mitigation: Streaming low-res preview during render; final-render handoff; 'cancel + partial refund'.

L:medI:high
#6
K12 signature verification is still TODO (real crypto, not format-only)

Mitigation: Precondition. Implement using @noble/curves (already a dep). Blocks Phase 4 entirely.

L:highI:crit
#7
Cold start — zero ANNs at launch means zero value

Mitigation: Seed with 5–8 production-quality ANNs before opening the marketplace.

L:medI:high
#8
Legal — output resembles a copyrighted style

Mitigation: Generation provenance attached to every artifact; takedown workflow via audit log.

L:lowI:high
Want to follow along?

Phase 0 is the documentation you just read. Phase 1 is the centralized prototype — a 30-second MP4 from a one-paragraph prompt, generated end to end by the Aigarth platform. Re-evaluation gate at the end of Phase 1.