humantic_content_personalize_v1

Rewrite an email you have already drafted, or write a cold first pitch from scratch — both matched to the recipient’s personality.

At a glance

Key inputs

responsetype (personalizedemail / firstpitch) + recipient id (email or linkedin_url) or personalitytype; optional subject, recipient_fname, sender_fname, force.

Upstream API

POST /v1/content-personalization

Two modes

personalizedemail

Personalizes an email you already wrote. Pass your draft as content. Your offer, your call to action and your signature stay intact — tone, structure, length and subject line are retuned to the recipient.

firstpitch

Writes a full cold email from scratch off a brief. Nothing to draft first. Both brief fields — prompt, user_context (5–1000 words each) — are required; leave one out and the call fails with INVALID_INPUT before it spends a credit.

Example ask

Here’s the email I drafted for Sarah at Acme — rewrite it for how she likes to be sold to, but keep my pilot offer and the Tuesday ask.

Example tool call

humantic_content_personalize_v1({
  "responsetype": "personalizedemail",
  "linkedin_url": "https://linkedin.com/in/sarah-chen-vp",
  "recipient_fname": "Sarah",
  "sender_fname": "Rohith",
  "subject": "Quick question about your prospect research",
  "content": "Hi Sarah,\n\nI noticed Acme is hiring 12 ops roles this quarter… We're offering a free 30-day pilot for teams of 10+ reps. Would you be open to a 20-minute call next Tuesday?\n\nBest regards,\nRohith Veerajappa\nHumantic AI"
})

Example output

{
  "status": "200",
  "message": "Success",
  "results": {
    "subject": "Prospect research as Acme grows",
    "content": "<html><body><div>Hi Sarah,</div><div><br></div><div>I noticed Acme is hiring 12 ops roles this quarter. As the team grows, keeping reps focused on conversations instead of pre-call research could be worth a look.</div><div><br></div><div>Humantic AI puts buyer intelligence directly inside the AI assistant your reps already use, so prep can drop from hours to seconds.</div><div><br></div><div>We're offering a free 30-day pilot for teams of 10+ reps, including onboarding support. Would you be open to a 20-minute call next Tuesday or Wednesday to walk through it?</div><div><br></div><div>Best regards,</div><div>Rohith Veerajappa</div><div>Humantic AI</div></body></html>",
    "meta": {
      "response_type": "personalizedemail",
      "parchtype": "DC",
      "pcolorgroup": "RED",
      "pbasedtone": "logical and descriptive using numbers when possible",
      "readinglevel": "5th grader",
      "emailwordcount": 100,
      "personalised_tip": [
        "The email references Acme hiring 12 operations roles this quarter, connecting growth to the need to protect sales reps' time."
      ],
      "profile_used": { "note": "the DISC type, tone rules and word-count target applied to the rewrite" }
    }
  },
  "metadata": { "status": "FOUND", "status_code": 1, "message": "The insights are fetched successfully." },
  "_status": "ready",
  "_idempotency_hit": false,
  "_trace_id": "dd2c513a72633558",
  "_quota": { "pool": "mcp_cp", "limit": 50, "consumed": 1, "remaining": 49 }
}
// content comes back as HTML; the offer, the CTA and the signature are carried over from the draft verbatim

Writing from scratch instead

humantic_content_personalize_v1({
  "responsetype": "firstpitch",
  "linkedin_url": "https://linkedin.com/in/sarah-chen-vp",
  "prompt": "Pain point: reps spend hours researching each prospect and still don't know how that buyer wants to be sold to. Value prop: buyer intelligence in seconds, inside the AI assistant reps already use",
  "user_context": "I'm Rohith from Humantic AI, a buyer-intelligence platform used by Sandler, Domo and IBM"
})

Good to know. Tone, length and framing are tuned to the recipient’s DISC type — same message, retuned for how they like to be sold to. It works without a stored profile (upstream resolves one, or pass personalitytype like "DC" directly), but profiling the person first with profile_create gives a higher-confidence rewrite. Each generation spends a content-personalization credit from its own pool (_quota.pool: "mcp_cp"); identical arguments within 24 hours return the cached result free — _idempotency_hit tells you which happened, and force: true regenerates.