AI brief foundry

Messy inputs in.
Executable briefs out.

For marketing teams, creative agencies, video studios, and brand teams. Client emails, Slack threads, and meeting notes become one structured brief — with sources cited and gaps flagged.

1 Submit 3–5 input packs 2 48h brief diagnosis 3 Full brief pack delivered
Apply for pilot View sample pack
Brief assembly · live Assembling · 78%
Client email · "need Q2 campaign ASAP" Slack · #client-acme · scope TBD Zoom transcript · 47 min kickoff Voice memo · 2:14 · "keep it under budget" Brand PDF · tone & mandatories Competitor landing page link
Structured brief · assembling
Goal

Drive Q2 pipeline for ops leaders via LinkedIn + landing page

Audience

Director-level ops · 200–2000 employee SaaS

Channels

LinkedIn paid · landing page · email nurture

Deliverables

60s video · 3 ad sizes · LP copy deck

Timeline

Creative lock Apr 15 · launch May 1

Open question

Budget ceiling not confirmed — follow-up required

Messy input wall

Six source types — studio whiteboard

Client emails

Goals · deadlines · stakeholders

View extraction →

Meeting notes

Decisions · action items

Meeting → brief →

Voice transcripts

Intent · tone · constraints

Voice → brief →

Slack discussions

Scope changes · quick calls

Brand materials

Tone · mandatories · visual rules

Brand memory →

Old project files

Past briefs · campaign learnings

Brief interrogation room

AI asks what humans forget to

Unclear goal

"Make something fresh for Q2"

→ What single metric defines success?

Vague audience

"Marketing leaders"

→ Primary segment + company size?

Incomplete deliverables

"Video and some social"

→ Formats, durations, platform specs?

Vague timeline

"Launch sometime in Q2"

→ Milestones: creative lock, media live, reporting cadence?

Missing questions module
Brief anatomy

Every slice links back to a source

Background

Kickoff context

↳ Zoom transcript · min 12
Goal

Pipeline SQL target

↳ Client email · Mar 1
Audience

Ops leaders · mid-market SaaS

↳ Strategy deck + Slack
Core message

Reduce ops chaos without new tools

↳ Voice memo · 2:14
Reasons to believe

Customer proof + benchmark stat

↳ Brand PDF · p.8
Brand tone

Direct, confident, no jargon

↳ Brand manual v4
Deliverables

Video · LP · 3 ad sizes

↳ SOW PDF
Channels

LinkedIn · landing page · email

↳ Slack #media
Constraints

No stock photos · legal review

↳ Brand mandatories
Reference materials

Competitor LP + past campaign

↳ Attachment links
Approvers

Client brand lead + agency CD

↳ Kickoff transcript
Deadline

Launch May 1 · lock Apr 15

↳ Client email
Open questions

Budget ceiling TBD

↳ AI gap analysis
Why briefs break

Margin notes — not a data wall

Brief quality

82% of clients · 79% of agencies

ANA creative brief research — 914 brands and agencies. A clear, tight brief is key to great creative work. Source: MediaPost / ANA

Rework cost

~33% of marketing budget wasted

IPA BetterBriefs — poor briefs and misdirection drive constant rebriefs. Source: ipa.co.uk

Time lost

~25% of time hunting answers

Atlassian State of Teams — 12,000 knowledge workers surveyed. Teams waste a quarter of the week searching for information. Source: atlassian.com

Creative director's desk

Before vs after the brief

Before · raw materials

  • Contradictory email threads gap
  • Unconfirmed budget in Slack conflict
  • Missing approver names
  • Scope drift mid-project
  • Unverified product claims

After · final brief

  • Single structured brief
  • Project goals + KPIs defined
  • Creative direction + tone locked
  • Delivery checklist by channel
  • Prioritized open questions for client
Brief type sample book

Eight formats — each its own page

CampaignBrand launchVideo productionLanding page Paid socialEmail campaignProduct announcementEvent
Open sample book
AI editorial workflow

Raw input → role-specific briefs

01

Process inputs

Emails, docs, transcripts, web pages

02

Extract facts

Goals, constraints, deliverables

03

Find contradictions

Flag gaps and conflicts

04

Generate questions

Supplemental client pack

05

Output versions

Copy · design · video · media · client approver

Multi-pass OpenAI pipeline — extraction, contradiction checks, multi-round questioning, and persona-based summaries. Not single-shot chat.

Brief quality score

Editorial manuscript stamp

87Brief clarity

Goal clarity · audience specificity · deliverable completeness · constraint clarity · brand tone consistency · approval path completeness · open questions — each with improvement notes.

See scoring module
Client follow-up pack

When the brief is incomplete — ranked by priority

Clients see the most critical gaps first — reducing rebrief cycles and back-and-forth.

Email draftClient formSlack messagePortal card
Follow-up pack →
Team handoff

One brief — six perspectives

Creative DirectorDesignerCopywriterVideoMediaClient approver
View handoff folders
Editorial submission

Submit 3–5 input packs

Not a product trial — a 48-hour brief diagnosis. Receive complete brief, missing-question list, quality score, team handoff versions, and client follow-up pack.

Outputs are internal working drafts — not legal, medical, or financial advice. Client approval required before production.

We reply within 2 business days. See sample deliverables first.