Consultant playbook
How to position, scope, and implement the AI Gateway so customers see clear value quickly.
The one-line positioning
"Your team already has the data. The AI Gateway lets them get answers by asking, in plain language, inside the permissions they already have, with a full record of every action."
Who to sell it to, and how to frame it
| Audience | What they care about | How to frame it |
|---|---|---|
| CEO / COO / CFO | Growth, risk, return | Faster decisions without new headcount or risk. Full accountability through the audit log. Lead with outcomes, not features. |
| VP Sales & Marketing | Speed, customer insight | Answers about customers, pipeline, and product performance in seconds, with no report backlog. |
| IT / Operations | Security, control | Permissions are enforced unchanged, data stays in the site, and every action is logged. Be precise and honest about how access works. |
| Finance / Store ops | Accuracy, ease | Reconciliation figures and stock positions on demand, in everyday language. It saves time and it is easy. |
Establishing the need
Open with the gap, not the technology. Ask the customer:
"When a store manager needs a number, how long does it take, and who do they ask?"
"How many of your staff actually know which report answers which question?"
"When someone changes a record, can you see who did it and when?"
Each answer points to the value: time lost to finding answers, training cost, and accountability gaps. The AI Gateway closes all three.
Scoping a first implementation
Start with reading and reporting. Enable the Core pack and let the team ask questions and run reports. This delivers value with zero risk to data.
Identify three frequent questions. Turn them into prompt templates so answers stay consistent.
Write one or two skills for the customer's specific routines, such as their month-end close.
Introduce write actions carefully, restricted to the roles that already perform them.
Reserve analysis and code for trusted analyst roles only.
Talking points that land
"It respects your permissions." The assistant can never see or change anything the user could not already see or change.
"Your data does not leave your site." The records stay where they are.
"Everything is on the record." The audit log captures every action against a named user.
"It uses your real reports." The numbers are the system's numbers, not an AI's guess.
"Start small, grow with confidence." Begin read-only, add capability as trust builds.
Handling cross-questioning
"Can the AI see data my staff should not?"
No. Every action runs under the signed-in user's own permissions. If they cannot open it in iVendNext, the assistant cannot reach it either.
"What if the AI makes a mistake on our data?"
Read questions never change anything. For changes, the user reviews what the assistant proposes before it acts, write tools can be restricted to specific roles, and every change is logged so it can be traced and reversed.
"Where does our data go?"
It stays in your iVendNext site. The assistant receives only the specific answer to the specific question asked, under your control over which assistant connects.
"Do the numbers match our reports?"
Yes. The assistant runs your existing reports, so the totals are produced by the same logic you already rely on.
"What can we control?"
Everything: which packs are on, which tools are enabled, which roles may use each tool, how long logs are kept, and who may connect.