Sai reads the source document in your Drive and pulls the segment from your Sheet, then writes a subject line with two alternates, a preheader, and a 400-word body with one CTA. The result is staged as a scheduled draft for the date you set. Nothing sends until you review it.
The recording is a real session. The sheet on the right is what it produced.
Sai opens each profile, pulls the signal, and writes the row, live, in a real browser.

Eight columns, sorted by score, with a source link behind every claim.
The audience description, the topic, and where the source material lives. A Sheet or contact list for the segment. Gmail or Mailchimp access.
A subject line with two alternates, a preheader, and a 400-word body with one CTA, with facts traced to your source material — plus the recipient count and how the segment was identified. Staged as a scheduled draft. Nothing is sent.
Monthly, so each issue starts with the source material already read and the segment already pulled.
A newsletter campaign is a single scheduled send to a subscriber list, consisting of the audience it goes to, the content it carries, the subject line that introduces it, and the time it is sent.
Building one involves two distinct kinds of work. One is production: deciding what the issue is about, selecting which material goes in, writing the copy, and writing the subject line. The other is assembly: loading that copy into an email platform, applying a template, selecting a list or segment, and scheduling the send.
Email platforms — Mailchimp, Beehiiv, Substack, Klaviyo, ConvertKit — handle the assembly side. They provide the editor, the template, the list management, the deliverability infrastructure and the reporting. What they do not provide is the decision about what the issue says, which remains with the sender.
Newsletter campaigns are also regulated. The FTC's CAN-SPAM compliance guide sets requirements that apply to commercial email regardless of platform: the subject line must accurately reflect the content of the message, the sender's header information must not be deceptive, the message must include a valid physical postal address, and opt-out requests must be honoured within 10 business days. The guide notes that each separate email in violation is subject to penalties of up to $53,088.
An email platform's editor opens to an empty template. The template solves layout. It does not solve what goes in it.
For a newsletter drawing on the sender's own material — product changes, published posts, customer conversations, research notes — the preceding work is reading through that material, deciding which items belong in this issue, grouping them into sections, and writing the connective copy. That work happens before the editor is opened, usually in a document, and it is the part that takes the time.
Generic AI newsletter generators approach this from the opposite direction. Given a topic, they produce plausible newsletter copy on that topic. The output reads as a newsletter and contains nothing the sender knows. For a company newsletter, where the value is in the specific updates only the sender has, topic-based generation reproduces the format without the content.
Founders and operators writing a company newsletter. The material exists across changelogs, published posts and internal updates. Assembling it into an issue is a recurring monthly task with no fixed process.
Content marketers running a regular send. The cadence is fixed and the constraint is per-issue production, not tooling.
Community and product managers. The audience is known, the updates are known, and the work is turning a period's worth of activity into something readable.
Writing it yourself scores full marks on the content columns. It is listed first because no automated method matches it on judgement about what belongs in an issue. Its constraint is that it recurs every month.
Sai reads the sources you point it at, and returns a complete draft issue: a section structure, the copy for each section, and several subject lines.
The sources are yours — a changelog, a set of published URLs, a document of notes, a folder of updates. The draft is assembled from what is in them rather than generated from a topic.
The output is a draft, not a send. It goes into the email platform, where the template, the list and the schedule are applied.
A monthly newsletter has a fixed shape and changing contents. The instruction — which sources, which audience, how many sections, what tone — is stable across issues. What changes is the material accumulated since the last one.
Running it on a schedule produces a draft on the same day each month, from whatever the sources contain at that point. The draft arrives before the editor is opened, which moves the monthly work from writing an issue to reviewing one.
The review step remains. A draft assembled from source material can misjudge which item leads, and that judgement is the sender's.
For the sending side rather than the drafting side, scheduling email in Gmail covers timed delivery from an inbox.
Where newsletter material comes from published long-form content, YouTube video summaries produces the source notes.
For the inbound direction — replies to a send that need structuring — email sender detail extraction handles that.