Cold outreach personalization you can cite.
One API call: a contact URL in, an evidence-backed opening line out — with the URL of the actual work it was derived from, so you can prove the research is real. Optionally, a verified personal email for the decision maker too.
Charged only on success. No result, no charge. No seats, no dashboard, no workflow builder.
One endpoint, one flag
url accepts a website or a LinkedIn profile URL. pitch is optional: with it, the opening angles toward what you sell; without it, you get a pure observational opener. Already have the email? Set need_email: false and pay only for the opening line. Need the email too? Flip the flag.
# you already have the email (from your CRM, Apollo, wherever) # $0.10 per successful call — just make the first line worth reading curl https://personalization.dev/v1/personalize \ -H "Authorization: Bearer pd_live_..." \ -H "Content-Type: application/json" \ -d '{ "url": "https://www.linkedin.com/in/sarah-chen-vp-eng", "pitch": "we detect flaky tests in CI and quarantine them automatically", "need_email": false }' # response { "status": "ok", "charged": true, "person": "Sarah Chen", "role": "VP Engineering, Meridian Labs", "opening": "Your QCon talk on cutting Meridian's CI queue from forty minutes to nine covered the retry-storm problem more honestly than most vendors do, especially the part about tests you can't delete but can't trust.", "evidence_url": "https://www.youtube.com/watch?v=example-qcon-talk", "observation": "Gave a QCon talk on reducing CI pipeline time from 40 to 9 minutes, including a section on managing untrustworthy-but-required test suites." }
# you only have a company domain or LinkedIn profile — need the email too # $0.50 per successful call: verified personal email + opening curl https://personalization.dev/v1/personalize \ -H "Authorization: Bearer pd_live_..." \ -H "Content-Type: application/json" \ -d '{ "url": "https://ovenandcrate.com", "pitch": "we run paid retention audits for DTC food brands doing 7 figures", "need_email": true }' # response additionally includes { "person": "Marco Reyes", "role": "Founder", "email": "marco@ovenandcrate.com", "email_source": "pattern_verified", "email_confidence": "verified" }
With need_email: true the email is part of the deliverable: if we can't find and verify a real person's address, the call is free. Generic inboxes (info@, hello@) are never returned.
Why evidence?
Every opening ships with the evidence_url it was derived from. Your compliance team can audit it. Your recipient can tell it's real. AI flattery gets deleted; a specific reference to something someone actually made gets replies. We never send email, we never invent facts, and generic role inboxes (info@, hello@) are never returned — a real person or no charge.
Pricing
| Call | Successful | Failed / no result |
|---|---|---|
| need_email: false — opening only | $0.10 | $0 |
| need_email: true — email + opening | $0.50 | $0 |
Usage-billed monthly. Check anytime: GET /v1/usage with your key.
Get a key
Early access is invite-only while we scale capacity. Email support@realtimecomms.co.uk with a line about what you're building.