Relationship Diagram

Livestream Production Airtable System

A series of interconnected bases for managing the production of a sales enablement livestream.

4bases
18tables
540fields
149computed
26relationships
core spine supporting relationship cross-base sync primary field ƒ ↑ Σ formula · lookup · rollup linked record Hover a table to trace its relationships · click for detail
EVENT PRODUCTION BASEone per event · schema deployed from centrally managed library

🗣 Speakers

44 flds
people featured in segments≈ owners & stakeholders
  • Name
  • ·Title
  • Lower Third Asset

🗺️ Locations

25 flds
venues & filming sitesvenues & filming sites
  • Location
  • 2026 Crew
  • Additional Assets

🎬 Crews

5 flds
camera / AV vendors syncedcamera / AV vendors synced
  • Crew Name
  • ·Booking Status

Elevated Slides

30 flds
custom motion graphics≈ creative request intake
  • Slide ID
  • Slide Type
  • Pre-Rec Video
  • Slide Content

Slide Types

6 flds
slide template catalog≈ creative template catalog
  • Name
  • ·Template + specs

🗓 Show Agenda

121 flds
the hub — one row per segment≈ campaign / wave definition
  • ƒSegment Title & ID
  • ƒSegment ID
  • ·Segment Type
  • ƒStart Time
  • ƒWeb App Session Link
  • ΣPre-Rec Video Status
  • + 114 more · 44 computed

🧑‍🤝‍🧑 Team

11 flds
staff → assignments everywhere synced≈ roles & access groups synced
  • Name
  • ·Email
  • ·Production Role

🎥 Pre-Recorded Videos

108 flds
one row per recording deliverable≈ channel execution row
  • ƒPre-Recorded Video
  • ·Video Status
  • ·Latest Draft Link
  • ΣSegment ID (from agenda)
  • ΣSpeakers
  • + 103 more · 37 computed

Agenda Team Resources

15 flds
join table: team member ↔ segment≈ per-channel assignments
  • Resource Name
  • Agenda Item · Assignee · Recording

🏃 Workstreams

8 flds
program areas≈ campaign calendar tracks
  • Name
  • Workstream Lead

🏁 Production Milestones

24 flds
tasks, self-nested≈ tasks & approvals
  • Milestone Name
  • Milestones (Nested)
  • ƒStatus

Assets

44 flds
mirror of project-specific assets in shared library synced≈ campaign links & creative synced
  • Asset Name
  • ·Status
  • ·Latest Draft Link
  • ·Latest File Name
  • ·Stream Link

Recording Studios

8 flds
links for producers to join remote recordingslinks for producers to join remote recordings
  • Studio Name
  • ·Studio URL

Streaming Channels

12 flds
live stream + chat endpoints≈ delivery endpoints
  • Channel Name
  • ·Playback URL

Livestream Edits

8 flds
any issues noticed mid-stream to be edited out in post-production≈ QA / change log
  • Issue
  • Segment · Submitter
ASSET LIBRARY BASEshared · all events

Assets

63 flds
primary asset records≈ shared link & creative library
  • ƒAsset Name
  • ƒFolder Name
  • Asset Type
  • ·Status

Asset Types

4 flds
type templates (L3, title card…)≈ link / creative types
  • Name
  • ƒFile Name Pattern

Keys are computed, never manually input

Every folder, filename, review link and notification in this system keys off one computed string. Parent records contribute their part; the formula assembles it; and it's properly referenced everywhere it needs to be.

This system — segment record key

informed by data rolled up from linked records

26_EVNT_PREC_R054_EV22279
  • 26_EVNT — event code, set once per production
  • PREC — deliverable type
  • R054 — recording
  • EV22279 — segment number, auto-assigned

Campaign CRM — tracked link

same shape: assembled once, upstream, by formula

utm_campaign=q3_launch_w2&utm_medium=rcs&utm_content=offer_a
  • q3_launch_w2 — from the campaign record
  • rcs — from the channel record
  • offer_a — from the link record

Campaign defines → channel inherits → link assembles. The string is a formula — so it is always accurate and up to date depending on the relevant variables.

Beyond the schema

Clean relationships are the foundation — Airtable layers working surfaces, automation and integrations on top of the same records.

Purpose-built UIs

Interfaces for every workflow

Each team works in an interface shaped to its job — producers in a run-of-show, designers in a render queue, reviewers in an approval view — all reading and writing the same records. No exports, no copies.

Native automation

Automations watch the tables

Status changes trigger the next step: a record meets conditions and a notification is sent; approvals happen, and the next task is created, with the right teammate assigned based on established logic. All within the base itself.

Platform integration

Connected to the tools around it

In this system, assets move from file storage to the video-review platform automatically, and every resulting link is written to Airtable — automations talk to Box and Slack while Airtable remains the source of truth.

Example Automation
An editor uploads a new draft to the Box folder — everything after that is automatic.
script runs notifies producer Box new draft uploaded Airtable link + status written Slack message sent
The Airtable update
ƒAsset Name ·Status ·Latest Draft Link ·Latest File Name
Example Segment | EV22279 Ready for Review https://app.box.com/s/9f3k2m1q8w7e5r4t6y0u 26_EVNT_PREC_R001_EV22279_Example_v1.mp4
The Slack notification
New Draft for Review3:42 PM
26_EVNT_PREC_R001_EV22279_Example_v1.mp4
Pre-Record Name: Example Segment
Post-Producer: @post.producer
Editor: @editor
Watch on Box Airtable

Why this architecture transfers

Isolation by design

One base per event, one shared library

Bases draw the isolation boundary — one per siloed team or workstream — and shared data syncs across them: one-way for reference directories, two-way where a team writes back. Every record keeps exactly one source of truth. Interfaces layer granular access on top: permission types and collaborator-tagged records decide who sees what, down to the row.

Computation Is Automatic and In-Base

Computed fields replace chained tabs

Timezone conversions, status roll-ups, character-count validation — all formula → lookup → rollup across linked records. The pattern of 30 Excel tabs, with sheets feeding sheets, becomes linked fields forming relationships between tables in a proper but accessible database.

Centrally managed schema

Managed apps & components, not copied Excel files

The schema is managed like software: a base like this ships as a managed app — built in a development base, published to the org library, updated centrally — so connected bases inherit changes instead of drifting. Standard building blocks (a UTM builder table, link fields) publish as locked components and stay identical everywhere they're used.