A webhook takes a LinkedIn URL or email. Perplexity pulls 90 days of company news, funding, and headcount shifts. Claude Sonnet scores against your ICP JSON and writes a 3-sentence opener referencing one specific signal. Smartlead queues the email at 9:45am local; Expandi fires the LinkedIn touch 36 hours later. End-to-end: under 3 minutes per prospect.
Tool-interchange map · every layer is swappable
What it does
You feed in a webhook and it comes out the other side as a sent LinkedIn message or email, personalised to the prospect, qualified against your ICP, and written in your brand voice. No manual research. No copy-paste. No delay.
Perplexity runs the research pass. Claude scores the lead and writes the opener. Expandi delivers it. The whole chain runs in under three minutes per prospect. The default stack is opinionated because it works. Every step has six alternatives if your preferred tool sits elsewhere.
This template is part of the Buttermilk Lead Engine and the Arcus Outreach System. Both run it at volume, daily.
Real-world use cases
B2B outbound agency
You buy a list of 500 leads on Monday. By Thursday a VA has researched maybe 40 of them.
Before
A junior researcher spends three days reading LinkedIn, news, and earnings calls before any message goes out. Most of the list never gets touched.
After
The webhook receives the list. Inside three minutes per lead, every prospect has a written-up dossier, a fit score, a personalised opener, and a queued send. The VA reviews edge cases instead of doing the research.
Solo founder selling to enterprise
You spotted a buyer signal on Twitter, want to send something today, and have 15 minutes between meetings.
Before
You open a tab, skim their site, draft a generic-sounding cold message, hit send, and feel weird about it.
After
You drop the LinkedIn URL into the trigger. By the time you finish your coffee, a researched, signal-aware first touch is sitting in your drafts ready to send.
RevOps inside a Series-B SaaS
Marketing runs a webinar. 300 sign-ups. Sales needs to know within an hour which 30 are worth a call.
Before
A spreadsheet gets passed around. By the time anyone has scored the list, the warm window is gone.
After
The form-submit webhook fires the pipeline. Each registrant is enriched, scored against the ICP, and flagged in HubSpot before the recording is even processed.
What changes for your team
Real runs, observed
Trigger
Apollo list of 250 SaaS founders, uploaded Tuesday 9am
What landed
250 enriched profiles, 78 marked A-tier, 78 personalised messages drafted and queued to Smartlead by 9:14am
Time
14 minutes end to end
Trigger
Single LinkedIn URL pasted into a Slack /lead command
What landed
A research summary, fit score, and ready-to-send opening line in the same Slack thread
Time
Under 90 seconds
Default tool stack
Sample prompts
Time + scope
2 to 3 days
To ship
Fixed
Pricing model
Included
Source JSON
14 days
Tweak window
Live in these engines
Related templates
A Notion watchlist holds 10 to 20 creators in your niche. Every 6 hours, Apify scrapes their latest posts. When one crosses your engagement threshold (default: 3x the creator's 30-day rolling median), every commenter is pulled. Clay enriches: email, company, role, company size. Claude scores each against your ICP JSON and returns tier A|B|C with a one-line reason. A-tier lands in HubSpot with the post URL and their comment as the signal field. A single viral post typically yields 4 to 12 qualified leads.
View →Template · 2 to 3 daysUpload a HubSpot view or Clay table. n8n checks deliverability (catch-all domain, role-based email, bounced-in-last-90-days) and drops the junk. Claude Sonnet writes a 3-step sequence per contact: step 1 is a 45-word opener referencing their most-recent signal. Step 2 lands at day 3, leads with proof. Step 3 lands at day 5, soft ask. Smartlead throttles sends across warmed-up inboxes so each domain sends under 40/day. Replies post back into HubSpot tagged positive | neutral | unsubscribe by Claude Haiku. A 500-contact sequence costs about $4 in model spend.
View →Template · 2 to 3 daysFireflies emits the transcript.ready webhook. n8n passes the transcript to Claude Sonnet with a strict JSON schema (decisions, actions with owners and due dates, blockers, follow-ups). Owners resolve by fuzzy-matching speaker names to your Notion people DB. Each action writes a row in the project page; Slack pings every assignee with a deep link. P95 latency call-ends to Slack-ping: 72 seconds.
View →Ready to ship this?
Book a call. I review your stack, confirm the fit, and we start the same week.