# Pipeline Enhancement Ideas — Pressure-Test & Prioritization

> Companion to [pipeline-enhancement-ideas.html](pipeline-enhancement-ideas.html).
> Scores the ideas by effort × impact × feasibility and recommends a build order.
> Brainstorm/prioritization, 2026-06-17. Directional — revisit once pilot data exists.

Matrix infographic: [pressure-test-matrix.svg](pressure-test-matrix.svg)

## Scoring

Impact = how much it moves the core loop (cataloging speed, review trust,
conversion, defensibility). Effort = build cost to a usable v1. Feasibility
class = what it depends on (deterministic / model / token-metered / Auth).

| # | Idea | Impact | Effort | Feasibility class | Verdict |
|---|---|---|---|---|---|
| 1 | **Voice cataloging** | High | Med | Needs model — local Whisper = free-capable | **Do first.** Best pain-per-dollar. |
| 2 | **Cross-run dedupe** | High | Med | Needs model — pHash free; embeddings cheap | **Do first.** Unlocks scale + recovery. |
| 3 | **PII / privacy scrubber** | Med-High | Med-High | Needs model (vision) | **Do, fast-follow.** Differentiator. |
| 4 | Confidence-tiered promotion | High | Low | Deterministic | **Sleeper pick** — ship before models. |
| 5 | Provenance timeline | Med | Low | Deterministic (data exists) | Quick win; trust + debugging. |
| 6 | Throughput / SLA dashboard | Med (agency) | Low-Med | Deterministic | Cheap; pair with agency push. |
| 7 | Comparable-evidence cards | High | Med | Deterministic surface + token deep | High-trust; phased. |
| 8 | Outcome feedback loop | High (compounding) | Med-High | Deterministic, needs sold-price volume | Defer until pilot data. |
| 9 | Auction-pipeline + channel export adapters | Highest value | High | File export earlier; API needs Auth + connectors | Phase-2; the reseller/auction payoff. [Spec](auction-outbound-integration-spec.md). |
| 10 | Photo decomposition / video-sweep | Med-High | High | Needs model (segmentation) | Defer; accuracy risk. |
| 11 | Active-learning review | Med | High | Needs model + data | Defer; needs edit-history corpus. |
| 12 | Authenticity detector | Med | High | Needs model | Defer; false-positive risk. |
| 13 | **Rooms / zones / pickup areas** | High (agency ops) | Low | Deterministic | **Do before heavy AI.** Estate-sale crews organize by physical space. |
| 14 | **Label batch workflow** | High | Med | Deterministic + print/export | **Do before checkout.** Labels are the bridge between floor and ledger. |
| 15 | **Team roles & permissions** | High (agency) | Med | Auth-gated | **Plan now.** Needed before real multi-crew pilots. |
| 16 | **Bulk edit / mass label / batch discount** | High | Low-Med | Deterministic | **Do early.** Boring throughput feature, high daily value. |
| 17 | **Client reports** | High (agency trust) | Med | Deterministic ledger | **Do with reconciliation.** Families/consignors need clean outputs. |

## Reading the 2×2

- **High impact / low effort / deterministic — do now, no model needed:**
  confidence-tiered promotion (4), provenance timeline (5), throughput
  dashboard (6). Ship keyless, stay green in CI, and together materially improve
  throughput and trust before spending on inference. **#4 is the most
  underrated** — pure policy logic over the existing promote path that
  multiplies reviewer throughput.
- **High impact / medium effort / free-capable model — the headline bets:**
  voice cataloging (1), cross-run dedupe (2). Both have a genuine free local
  path (Whisper, perceptual hash), so they fit the cost-controlled rule and only
  escalate to paid inference when quality demands. This is where the "kills the
  pain" story lives.
- **Highest value but mostly Auth-gated — stage it:** auction-pipeline + channel
  export adapters (9). The literal "catalog once, list everywhere" promise and
  the reseller/auction pitch. The *file* exports (LiveAuctioneers / Proxibid /
  Auction Flex spreadsheet + lot-numbered image bundle) can ship without
  third-party Auth; the API + sale-results pull-back need Auth + connectors. See
  the [auction-outbound integration spec](auction-outbound-integration-spec.md).
- **Defer on feasibility risk, not value:** decomposition/video-sweep (10),
  active-learning (11), authenticity (12). All need models *and* either accuracy
  guarantees or a training corpus not yet available, with no free fallback.

## Recommended sequence

1. **Deterministic trio** — confidence-tiered promotion → provenance timeline →
   throughput dashboard. Fast, keyless, de-risks the queue.
2. **Voice cataloging + cross-run dedupe** — the two free-capable model wins that
   carry the marketing story.
3. **PII scrubber** — fast-follow trust differentiator.
4. **Comparable-evidence cards**, then **outcome feedback loop** once sold-price
   data accumulates.
5. **Auction-pipeline + channel export adapters** — file exports first (no Auth),
   then the API + sale-results pull-back when Auth + connectors are live.

**Trap to avoid:** the flashiest ideas (video-sweep, decomposition,
authenticity) are the *worst* first bets — model-dependent, no free fallback,
quality-sensitive. The deterministic policy work is unglamorous but it's where
the cheap throughput gains actually are.

## EstateSail / POS Lesson

EstateSail shows that estate-sale companies care about the whole operating
loop: rooms, labels, bulk edits, team roles, customer check-in, checkout, and
client reports. We should absorb those operational basics, but not build a full
POS first.

Update the build order before heavy AI:

1. **Rooms / zones / pickup areas** - attach physical location to capture,
   labels, pickup, reports, and reconciliation.
2. **Label batch workflow** - choose items, group by room/status/value, preview,
   print, reprint, or bind sticker-pack codes.
3. **Bulk operations** - bulk room/category/price edits, mass labels, batch
   discounts, duplicate room setup between sales.
4. **POS interop + client reports** - import EstateSail/PROSALE/Square/Shopify
   outputs into `inventory_status_events`; generate family/consignor reports.
5. **Team roles & permissions** - admin, cataloger, pricer, label/runner,
   checkout, and report viewer.

The POS stance: **interop first, replacement later only if earned**. Let sellers
start with EstateSail, PROSALE, Square, Shopify, or manual checkout. TroveSnap
uses those systems as onboarding bridges, then gradually pulls the important
truth into its own identity, demand, and recovery model.

## Source documents

- Ideas catalog: [pipeline-enhancement-ideas.html](pipeline-enhancement-ideas.html) / [.svg](pipeline-enhancement-ideas.svg)
- Build brief: [codex-handoff-freemium-positioning.html](codex-handoff-freemium-positioning.html) (§ Phase-2 pipeline sequence)
- Folder index: [index.html](index.html)