AIAP Tool Guides·Humanic AI·24:46

Humanic Deep Dive: Prompt-First Email Marketing - Multi-Email Sequences, Contacts and Cohorts, Gmail Sending, and a Five-Day Course

Akshat (surname unclear) Solo presenter using the tools as an outside user; demo brand "Better With AI"

The short version

Humanic starts from a prompt: "send a newsletter regarding how AI can help you grow in career" became three emails timed now, +3 and +6 days, with variations, mobile preview, code view and spam insights (0:00-0:06).
Contacts by hand or CSV, cohorts as segments, sending through a connected Gmail; the test email arrived with an unsubscribe link "in less than five minutes" (0:02-0:11).
Cold outreach with brand documents, model choice (Sonnet or Opus) and voice answers; a five-day paid email course generated as one connected set and edited across all emails at once (0:12-0:24).

At a glance, three clicks deep

Skim here first: the closed row is the glance, open is the study card with the key points and timestamps, and the ↓ link drops to that concept's full write-up below.

01Prompt to campaign: a whole sequence with timingDescribe the campaign;›

Describe the campaign; Humanic asks follow-ups, writes every email with send delays, and lets you edit by instruction.

First campaign (0:01-0:06)

Course set (0:16-0:22)

Checks (0:22-0:24)

↓ Full write-up of this concept

02Contacts, cohorts and the sending identityLoad contacts (CSV needs an email column), group them into cohorts, and send from a connected Gmail or a ve…›

Load contacts (CSV needs an email column), group them into cohorts, and send from a connected Gmail or a verified domain.

Contacts (0:02-0:04)

Sending (0:07-0:11)

Domain settings (0:15)

↓ Full write-up of this concept

The concepts in full

Prompt to campaign: a whole sequence with timing

how-to

Writing the set together keeps "coming tomorrow" references consistent.

Edits by prompt ("reduce the content in day six email and keep everything as it is"; "make sure the author on each of these emails is written as ...") apply where asked or across the set. Insights flag spam triggers ("free", "millions", "billions"). The five-day course prompt covered AI basics, prompt frameworks, tool differences (ChatGPT, Claude, Gemini, NotebookLM), advanced Claude use and AI at work, with design instructions (Material style, Poppins, red highlights). Check with a test send, mobile preview, code mode and spam check; download HTML.

Do it in this order

01Prompt the campaign (audience, goal, number of emails).
02Answer clarifying questions; review emails and delays.
03Edit by instruction; check insights and mobile preview.
04Send a test to yourself before any real send.

Why it matters

Sequences in minutes instead of days.

Contacts, cohorts and the sending identity

how-to

Do not send the same email to everyone - that is what cohorts are for.

He connected Gmail by clicking through Google's unverified-app warning to allow sending, created a cohort of his own two addresses, set a sender name and activated. Custom domains need DNS records at the registrar (GoDaddy, Namecheap) - shown, not configured; deliverability, limits and pricing are not covered, and all sends went to himself on dummy data. Before cold outreach at volume, set up domain authentication separately.

Do it in this order

01Contacts > add or import CSV.
02Create a cohort (manually or by asking in chat).
03Connect a sender (Gmail or a domain with DNS records).
04Name the campaign, set sender name, activate; check Email logs.

Why it matters

The parts that decide whether email actually lands.

This idea elsewhereparallelCatalyst Session 7: anatomy of a cold email

Tools referenced

Tool	Coverage	Moment	Context
Humanic	demonstrated
Gmail	demonstrated
PostHog	mentioned
Wispr Flow	mentioned
Claude Sonnet 4	mentioned
Shopify	mentioned

Session materials

Archived locally on V: — click to open. Companion pages link to the LMS.

lms

Tool Deepdive

companion page on the platform

lms

Tool Deepdive

companion page on the platform

lms

Tool Deepdive

companion page on the platform

lms

Tool Deepdive

companion page on the platform

lms

Tool Deepdive

companion page on the platform

Action items

Resources mentioned

Resources

docLesson page on the platform

Extraction notes

This page was built from an auto-generated transcript, which garbles product and people's names. Those were corrected silently in everything above and logged here for transparency. The warnings flag claims that were true on the recording day but change fast.

Transcript corrections applied

The transcript says	The trainer actually means
umani / Humanix / the money	Humanic
post hoc	PostHog
Whisperflow	Wispr Flow
cloud co-work or cloud code	Claude Cowork or Claude Code

True on recording day — verify before relying

Vendor UI walkthrough: menu names, layout, plan limits and pricing change monthly - check the live product before following click paths.
Model names (Sonnet, Opus), the Google warning click-through and the integrations list are as of recording; pricing and send limits were not covered (0:07-0:15).