Blog/Private AI as an executive assistant for scheduling and correspondence
Private AI

Private AI as an executive assistant for scheduling and correspondence

By Samuel Seidel · September 9, 2026 · 8 min read

A model on a dedicated machine can do the language part of executive-assistant work, turning "find 45 minutes with the board chair before the 14th, avoid Tuesdays" into proposed times and a polite email, without a cloud vendor reading the calendar. Cloud scheduling assistants do this well, but they need read access to calendar and inbox, so every meeting title, attendee list and thread passes through the vendor. For a calendar with board prep, M&A diligence calls or compensation discussions, that's a lot of exposure for a drafting task.

What a calendar gives away

A calendar entry titled "Board comp committee: CEO pay review" or "Call w/ [acquirer] counsel re: diligence" tells a reader a lot even without opening the invite. A scheduling assistant that ingests a full calendar to find open slots and draft invites necessarily sees all of that, plus the attendee list, which can itself reveal who's involved in a still-confidential process before it's meant to be known. Correspondence is similar: a thread summarized or drafted around a personnel matter or an unannounced deal carries the same risk as the document itself, because the AI tool now has a copy.

None of this means cloud scheduling tools are reckless. It means the tradeoff is worth making deliberately for a general assistant, and worth reconsidering specifically for the subset of an executive's calendar and correspondence that touches board matters, M&A, or anything under an NDA.

What self-hosting changes

Running the assistant model on a dedicated DGX Spark means calendar data and draft correspondence go to your own model on a machine assigned only to you, not through a third-party scheduling product. You keep your existing calendar and mail platform; only the AI component that reads them to draft invites and messages moves. The Spark is a GPUwerk-run machine (its own in Czechia, or a partner's in the country the console shows), operated under its published DPA. Building the integration against a test calendar and mailbox first keeps early mistakes out of the executive's real accounts.

Where it helps

The strongest use is turning unstructured input into structured drafts: a voice memo becomes a meeting invite with a proposed agenda, or bullet points become an email in the executive's usual tone. It also handles first-pass triage, summarizing a long thread in three sentences so the executive can decide whether to read it, and routine scheduling replies ("happy to do Thursday, does 2pm work") that a human reviews before sending.

It's a poor fit for anything requiring real judgment about relationships or politics, like deciding how to phrase a delicate decline to a board member's request, or reading between the lines of a terse reply from an acquirer's counsel. Those calls belong with the assistant or executive who knows the relationship, not with a model pattern-matching on email tone.

Quality tradeoffs

For structured drafting (notes into invites, thread summaries, routine replies) a self-hosted model does the job and the output needs light editing. It's noticeably weaker at matching a specific executive's voice on nuanced or high-stakes correspondence without a fair amount of example text to work from, and it won't pick up on office politics or relationship history the way a longtime human EA does. Expect to keep a human in the loop on anything that goes out under the executive's name to a board member, investor, or counterparty, which is good practice regardless of which AI is drafting the first pass.

Setup effort

Most of the setup is connecting the model to read-access views of the calendar and inbox it needs to draft from, which typically takes an integration with your existing calendar and mail platform's API rather than a wholesale platform migration. Give it a handful of example emails in the executive's actual voice before asking it to draft correspondence; without that, its default tone will read as generic. Plan for a short calibration period where the assistant or chief of staff corrects draft tone and phrasing until it settles into something usable.

One Spark covers an executive's office

This workload is lighter than document-heavy ones: mostly short drafts and summaries, not long-context reasoning. A single Spark at $0.79/hour has 128 GB of unified memory for the model plus the retrieval layer that reads calendar and email context, enough for one executive's office. Because the machine is dedicated, no other customer's requests compete with the scheduling back-and-forth.

A note on scope

None of this argues for running a full-blown personal assistant with unrestricted access to everything the executive touches. A narrower scope, calendar and correspondence around board and deal work, gets most of the benefit with less exposure than an assistant granted the whole inbox and calendar by default. Decide what the assistant reads instead of accepting the broad access a setup wizard offers.

Related pages

First top-up: pay $20, get $30 in credit

Draft executive scheduling without a cloud assistant reading the calendar.

A dedicated DGX Spark in EU-Central, $0.79/hour, for executive scheduling and correspondence drafting on a dedicated machine assigned only to you.

Your data stays on a machine assigned only to you. GPUwerk processes it under its published DPA.

Create account Private LLM hosting