All workMarketing Efficiency

AI Outreach Automation

Outreach & Crawler for FRM

AI Brand Discovery & Outreach

We built an AI outreach engine for a Melbourne talent agency. It finds fashion brands, matches each one to the right models and drafts the pitch for a person to send.

ClientFRM
IndustryTalent & modelling
BuildCustom web app
ServiceMarketing Efficiency
The FRM outreach dashboard, blurred: an inbox scan summary with replies found, bounces and clients warmed
The Problem

Cold outreach that couldn't scale.

FRM's growth ran on pitching fashion brands. Doing it by hand ate whole days, and off-the-shelf tools know nothing about models or comp cards.

  • Finding brands worth pitching meant hours of manual research.
  • Every pitch needed a comp card built by hand and an email written from scratch.
  • Matching the right model to each brand's aesthetic was guesswork under time pressure.
  • Outreach went out in fits and starts, with no view of what was actually landing.
  • Generic outreach tools don't understand models, comp cards or brand fit.

The part we insisted on: a person approves every send. We build outreach that drafts and never outreach that fires on its own, because the cost of a bad automated pitch lands on the client relationship rather than on us. FRM reviews and sends. The system does the research and the writing, which is where the hours actually went.

The Workflow

AI does the legwork. A human sends.

7 stages
Discovery to send
AI handles the research, matching and drafting. A person approves every single send. No brand ever gets an email FRM hasn't signed off on.
What It Does

The pipeline, step by step.

1

Discover brands

Fashion brands go into a cold pool (a list of brands nobody has contacted yet), ready to be analysed and pitched.

2

Analyse the fit

Claude reads each brand's website and works out its aesthetic, audience and the kind of model it books.

3

Match models

For each campaign, Claude pairs brands with the right models from FRM's roster by aesthetic and suitability.

4

Draft the pitch

Claude writes a personalised email for every match, and the renderer builds a PDF comp card for that specific model, with their stats and portfolio.

5

Review

Every draft lands in a review queue where the team edits, approves or rejects it. Nothing sends unreviewed.

6

Send & track

Approved pitches go out through Klaviyo with the comp card attached, and the dashboard tracks brands, sends and reply rate.

The Results

What changed.

Hours → minutes
Research, comp cards & drafting per campaign
Every pitch
Personalised, email + comp card per brand
100%
Human-reviewed before anything sends
Built Right

Fast, but never on autopilot.

A human on every send

AI drafts and a person at FRM approves. Nothing reaches a brand without being reviewed first, so a bad draft never costs them a relationship.

Personalised, not spun

Each pitch is written for the brand's actual aesthetic and the matched model, with a real comp card, not a mail-merge.

Their data, their stack

Runs on FRM's own Supabase, with brands, models, drafts and sends all in one place they own. No per-seat outreach tool.

Built to grow

Built so automated brand discovery (Apollo/Apify) and higher-fidelity comp cards can be bolted on if the volume calls for it.

The Stack

What we built it with.

Next.js TypeScript Supabase Claude API n8n Express pdf-lib Klaviyo
In Their Words

What the client said.

“We were losing whole days to brand research, comp cards and writing pitches one at a time. Now the system finds the brands, matches the right models and drafts the entire pitch, comp card and all. We just review and send. It's turned outreach from a job nobody had time for into something that actually runs.”
FRM

Want something like this built?

A couple of lines is enough. It goes straight to Aidan, not a sales team, and if it isn’t worth building he’ll tell you.

  • You deal with Aidan directly, not a sales team.
  • A straight answer on whether it's worth building, and what it'd cost.
  • You own everything we build. No lock-in and no per-seat fees.
Aidan Lambert, founder of Better Automations
FOUNDER
Aidan Lambert
He reads these himself, and he'll write back.