The offer

Four weeks.
Your numbers.

Pick one production; a wrapped one is perfect. Hand us its ShotGrid database or its screenplay and we turn it into the production's context brain, index the archive read-only, then run a context-aware test ingestion pipeline against it, with your agents on top. Nothing migrated. Nothing replaced. You keep everything we build.

The pilot schedule

The schedule
WEEK 0

Scoping, 60 minutes

Pick the production and the runway. We agree the asset set, success gates, access, and the sandbox for test output.

WEEK 1

The brain

Your ShotGrid export or screenplay becomes a queryable context brain: scenes, shots, cast and versions in one graph.

WEEK 2

The index

We index the production's archive read-only: checksums on every file, previews for RAW, EXR and legacy 3D, duplicates surfaced. Baseline measured.

WEEK 3

The pipeline

A context-aware test ingestion runs against the brain: files classified, tagged and routed, writing only to a sandbox you name. Sources untouched.

WEEK 4

The second unit

Agents go on top over MCP, the open agent-connector standard: questions that join ShotGrid and the script, on-model generation tests, the pipeline on demand. We walk the findings report with your team.

Two runways

Two runways · your call at scoping

Runway A · Our cloud (recommended)

Fastest start: nothing to install, no deployment to approve. The pilot runs on Fortify's existing cloud, pointed at archival data you choose: read-only credentials to an archive location, a copy-drop, or just a StorageDNA catalog export, which is metadata only. Week 1 starts within days. Single-tenant, encrypted, deleted and verified at wrap.

Runway B · Your environment

The production posture from day one: services in your AWS VPC or on an on-prem VM, compute and content inside your walls, with week-4 agent and generation calls only to providers you approve. The trade: security review and deployment approval come first, typically adding weeks before week 1.

What we measure, filled in ink
Time-to-find, one assetDuplicates reclaimedBlack boxes openedAuto-tagged by contextRe-makes avoidedRetrieval timeTD hours returned

No canned ROI. Every measure lands in the findings report in your numbers, not ours.

Questions
How long is the Asset Foundry pilot?

The Asset Foundry pilot runs four weeks, plus a 60-minute scoping session. Week one builds a context brain from your ShotGrid database or screenplay, week two indexes the production's archive read-only, week three runs a context-aware test ingestion pipeline into a sandbox you name, and week four puts agents on top and delivers a findings report in your numbers.

Does the pilot require installing software in our environment?

No. The fastest runway runs from Fortify's cloud, pointed at archival data you choose: read-only credentials, a copy-drop, or a StorageDNA catalog export that is metadata only. Deployment into your AWS or on-prem environment is the production posture and is available as the second runway.

What do we keep after the pilot?

Everything built during the Asset Foundry pilot: the context graph of your production, the searchable metadata index, and the findings report with every measure in your numbers.

The ask is small

Name the production.
We bring the plan.

Your commitment: the production, its ShotGrid export or script, and 60 minutes for scoping. Your exposure: archival data you choose, read-only sources, sandboxed output, deletion verified at wrap.