What Can Sales and Account Teams Hand to an AI Agent First?
Each build is one you can set up: what the AI agent reads, what it drafts, and what a person approves before anything reaches a customer.
Sales and account teams can hand an AI agent the preparation around a customer: a call brief, a discovery recap, a competitor comparison from published pages, a sort of quiet deals and a quarterly review agenda. In each published build a person edits and sends. These are builds you can set up.
BNEDai has published a step-by-step guide for each build below, with the instruction block to paste and a check to run on the first few uses. This page lists them for a sales or account lead deciding where a first pilot could start.
A good first sales task recurs and ends in a draft
Three things. The task recurs, so a draft is worth the setup. The output is a draft rather than an action, so nothing reaches a customer without a person. And the person who approves it already knows what good work looks like, so they can tell when the draft is wrong.
Sales work has plenty of this. A call brief is read before a call. A recap is edited before it is sent. A deal list is sorted by someone who knows which accounts matter. In every build below the instructions keep the person in that seat.
What the five builds share
Four habits, written into the instruction blocks. The AI agent reads only the sources it is given, such as one sheet, one set of notes or a list of pages. Each line it writes points back to where it came from, usually a quote. Where it cannot find something, it writes that, such as "not in the notes" or "not on their site", instead of filling the gap. And it sends nothing: the instructions say to send nothing or contact no one, and the meeting brief runs read-only, which blocks sending.
A brief before the customer call
You decide when automatic prep turns on, and only after you have checked one brief against what you know about that customer and noted what it missed. The meeting brief guide uses an AI agent that reads your calendar entry and the apps you connect in a read-only run. It hands back who is on the call, the prior context, the questions the customer has asked that you have not answered, and a few talking points tied to those items. Where there is no record it writes "Nothing found."
A recap and proposal outline after discovery
You edit the follow-up and send it yourself. In the discovery follow-up guide the AI agent reads the call notes you paste and drafts a short email to the buyer and a first proposal outline. It quotes the buyer only where your notes put words in quotation marks, lists the open questions with an owner and a date, and fills your own proposal headings, marking "needs my input" where the notes do not support a heading. It states no fee.
A competitor comparison from their own pages
You open the lines yourself before sellers see the comparison, and how you sell against anyone stays your call. The competitor comparison guide has the AI agent read only the pages you name, and no others, and return one page of rows. Each row quotes the competitor's line with its link and the date it was read, and says "not on their site" where a point is not published. It states a cost only when the competitor's own page prints one.
Quiet deals sorted by your rules
You read every draft and send it yourself, and the instructions say never to change the sheet. The quiet deals guide has an AI agent read your pipeline sheet, and nothing on the web, and place each quiet deal by rules you write into one of three piles: reach out now, wait until a date, or close it out. It shows the rule behind each placement and drafts a short note only for the first pile, in your own words.
A quarterly review agenda built around one ask
You decide what goes to the customer. In the quarterly review guide you state the one thing the customer should agree to, and the AI agent reads your notes and the account's sheet, sorts what it finds into what supports the ask, what works against it and the likely first topic, and drafts the agenda with the ask last. It does not invent results, and it lists anything missing under "find before the meeting."
Start with the meeting brief
Start with the meeting brief guide: it runs read-only, and it has you check one brief against what you know before turning on automatic prep. The pilot playbook covers what to write down first, and what an AI agent can touch covers the permissions. The AI Readiness Index is a free 18-question check on how your team runs AI today.
Frequently asked questions
What should a sales team hand to an AI agent first?
A recurring preparation task whose output is a draft, such as a call brief or a discovery recap, that a person edits before anything reaches a customer.
Do these builds send anything to customers?
In the published builds the instructions tell the AI agent to send nothing or contact no one, and the meeting brief runs read-only, which blocks sending. A person edits and sends.
Do these pages say what results a team got?
No. They describe how each build is set up, what it reads, what it drafts and what a person approves.
Reviewed on November 20, 2026.