VFX Pipeline Consolidation.

A VFX studio running on ShotGrid, spreadsheets, email, and Slack. Coordinators were the integration layer. New hires took two weeks to ramp because institutional knowledge lived in five places. Background inside the stack made the failure modes legible from day one.

4 → 1Platforms unified
14 → 3Days to productive (new hires)
100%Pipeline visibility

The problem

Every VFX studio runs the same fragmented stack. ShotGrid for shots. Spreadsheet for bids. Email for client comms. Slack for internal. None of it connected. Coordinators were the glue.

Visible cost: two weeks of ramp per new coordinator. Hidden cost: every shot that slipped through the gaps because the system could not see itself end-to-end. On a $100M production, a missed shot is not a coordination problem. It is a margin event.

The approach

One unified pipeline view. An AI onboarding agent in Slack, trained on internal documentation under tightly scoped access controls, answering the fifty questions every new hire asks in their first two weeks. Institutional knowledge becomes a queryable system. Engineered so the wrong question does not return the wrong answer to the wrong person, and so a mis-tagged shot cannot pollute downstream automation.

Before / after BEFORE: ShotGrid (shots) ┐ Spreadsheet (bids) │ Email (client comms) ├── Coordinator (manual integration layer) Slack (internal) │ Whiteboard (thumbnails) ┘ AFTER: ShotGrid (shots) ─────────────┐ Spreadsheet (bids) ───────────┤ Email (client comms) ─────────┼──► [Unified Pipeline Layer] Slack (internal) ─────────────┘ │ ▼ [Coordinator dashboard] + [AI Onboarding Agent in Slack]

The build

Four-week engagement. Integration-heavy because source data already lived in mature systems. The work was making them visible together for the first time without corrupting the upstream sources.

The outcome

New-hire ramp cut from 14 days to 3. Coordinators recovered hours per week. Institutional knowledge that lived in three senior heads now lives in a queryable system every coordinator has on day one. The studio no longer runs on memory.

Background context: $100M+ production work on Skyfall, Avengers, Game of Thrones. The failure modes of fragmented VFX pipelines compound across every shot, every milestone, every billing cycle. Staying outdated outruns the cost of the build.

Tech stack

ShotGrid API, Google Sheets API, Gmail API, Slack Bolt, PostgreSQL, Claude API (onboarding agent), Python, Next.js (coordinator dashboard). T-shape engineering: software on the vertical, operations on the horizontal. Four weeks. Built solo.

Working in VFX, gaming, or post-production?

Industry background built in. Discovery cuts straight to the operational problem.

Book a discovery call