Nobody funds a use case they cannot describe. A leader handed an AI mandate needs three artifacts before a budget conversation is possible: a short list of opportunities that fit the business, an architecture that carries them, and a number a finance leader will accept. Assembling those by hand is a professional-services project. Iternal built the AI Blueprint Builder to assemble them from an interview, and executives ask the plain question first: what is it, and what comes out?
AI Blueprint Builder
What Is the AI Blueprint Builder
and What Does It Produce?
A guided interview in two tracks, the deliverables it returns, how far its recommendations can be narrowed, and the finishing work you still own.
Built from real buyer questions in our sales meetings
Direct Answer
A guided interview that writes consulting deliverables. The AI Blueprint Builder is a chat-style interview that turns answers about your business into a scoped set of AI use cases and the written deliverables behind them. Iternal runs it in two tracks: a general use case exploration that surfaces and stack-ranks candidates, and a longer deep dive that builds one candidate out end to end. Iternal publishes the set as up to eight executive-ready deliverables — an executive summary, a forwardable buy-in brief, a solution recommendation, a cost estimation for finance, an implementation roadmap, a viability assessment, a stakeholder alignment map, and the strategic add-ons that carry governance, the ROI business case, a change and training plan and the pilot charter — and each completed blueprint carries a chat you can interrogate.
The limit: you receive a first draft that a person finishes. Iternal states plainly that the reports are never claimed to be perfect and need proofreading before they reach a client. Cost scoping returns a range rather than a definitive quote. The interview cannot be completed by someone who does not already know the customer and how a large organization is structured, and it leaves gaps that need further clarification. The finished set also takes real time to digest. Budget the finishing work.
Ask for today's numbers rather than trusting a fixed one. Session length, deliverable count and page count all move with the track you run, how prepared the answerer is, and how much Iternal has added since the last release. Read published figures as ranges with both ends intact, ask for a full sample deliverable set before you commit anyone's calendar, and get the current list in writing; Iternal answers the targeted questions below.
The tool and the method are separate things. The guided-discovery software is one of them: what the interview asks, what it writes, and what its recommendations can be restricted to. The method a team runs for itself is the other. For more information visit the use-case selection page.
Two Tracks: A Short Exploration, Then a Deep Dive
The interview adapts. Each next question is synthesized from your last answer rather than read off a fixed script, which is why two sessions on the same afternoon look nothing alike. Iternal splits it into two tracks, so a buyer can spend twenty minutes finding out whether AI applies before committing an afternoon to one opportunity:
| Track | What it asks | What it returns | Commonly takes |
|---|---|---|---|
| General use case exploration | Around twenty adaptive questions about the business, its goals and its workflows. | A report identifying and stack-ranking candidate use cases, plus a downloadable executive summary. | 15 to 30 minutes. |
| Detailed deep dive | Around a hundred questions, run in staged steps against one chosen use case. | The full deliverable set for that use case: architecture, governance, ROI and pilot scope. | 45 to 90 minutes, longer when the process is complex. |
Every recommended use case in the exploration report carries a button that opens the deep-dive interview for that opportunity, so the second track starts from the answers given in the first. A buyer who can already articulate the business can skip the exploration entirely. Teams still weighing whether their data, staffing and governance are ready for either track can gauge that first. For more information visit the AI Readiness Assessment page.
The clock tracks preparedness, not the software. Iternal describes a technical person who knows their own business finishing a deep dive in roughly 30 to 45 minutes, a first-time non-technical answerer commonly running 60 to 90 minutes, and a complex manufacturing process pushing a session well past the time it was given. Sessions running an hour or more against an expected 45 minutes are common enough that Iternal plans for them. Block the longer end and finish early.
What Comes Out of a Completed Blueprint
A completed deep dive returns a stack of documents rather than one report. Iternal publishes the set as up to eight executive-ready deliverables, the last of which — strategic add-ons — itself breaks out into governance, the ROI business case, a change plan, a training plan, the pilot charter and compliance depth. Where you land inside that depends on the track you ran and how completely you answered: the tool withholds some until you close the gaps it identified, each carries its own unlock, and Iternal keeps adding to the set. Ask for the current list rather than working from a count.
Page counts move too. Individual deliverables commonly run 10 to 20 pages each, and a full set is commonly described at around a hundred pages. Do not derive either figure from the other: how many of the eight you receive depends on how completely you answered, so the multiplication misleads procurement. If length matters, ask for a current sample set.
The output is built to be worked on. Every deliverable downloads and edits as a Word document, and the set also exports as plain text, markdown, HTML and PDF. Each finished blueprint carries a chat button as well, so a reader can ask the underlying data a question instead of reading a hundred pages.
Pin it down: questions for your evaluation
-
What is the current deliverable list for the track we intend to run, and can we read a full sample set first?The exact artifact you receive on the day you buy, judged by reading one rather than by a count that moves.
-
How much calendar time should we block, given whether our subject-matter expert has run one before?A realistic hold, because interview length tracks the preparedness of the answerer and the complexity of the work.
-
Which sections will need our own experts to finish, and what gets proofread before delivery?Where the finishing work sits, so you can staff it up front instead of meeting it at review.
Where the Recommendations Come From
Software is only as good as the expertise poured into it. Iternal names three sources behind the AI Blueprint Builder: two published books by its founder, one on enterprise AI strategy and one on the partner motion; roughly a thousand pages of research written for those books and cut from them; and thousands of AI consultations and implementations Iternal has run with customers. Behind the interface, Iternal states there are about 3 million tokens of custom prompting and a set of AI agents working in concert, which is what lets each next question follow from your last answer.
Two consequences follow. The output is generated rather than assembled from templates: Iternal states each blueprint is custom to that customer, industry and use case with no boilerplate, so the same interview run twice will not return identical pages. The depth of that corpus is an Iternal claim: ask for it in writing alongside the sample deliverables.
Agnostic by Default, Restrictable on Purpose
Iternal developed the AI Blueprint Builder product-agnostic. It takes the best of what is available — open source, commercial, or the brownfield systems you already run — and names Iternal software such as AirgapAI and Blockify only where they fit the use case. Where nothing in the portfolio fits, the output scopes the requirement as custom development, which is why a blueprint can recommend a cloud service Iternal does not sell.
Agnostic is the default, and the default is configurable. Iternal states the tool can be scoped by organization and by logged-in user: an account can carry a directive to recommend only one hardware line, to stay inside a chosen catalog, to recommend cloud platforms only, or to skip cloud entirely. Partners run branded tenants on exactly that basis. Both settings ship, so the practical question is which one governs the copy in front of you:
Pin it down: questions for your evaluation
-
Is the instance we will use running the agnostic default, or is it scoped to a particular catalog?Whether the recommendations you receive were open to the whole market or bounded before the interview began.
-
When a requirement falls outside the configured catalog, what does the output say?Whether a restricted instance still discloses the gap, or recommends the nearest thing it is allowed to name.
A Sample Table of Contents
Descriptions of consulting output are cheap; contents pages are not. The set below is the one Iternal publishes for a completed blueprint, with the audience each document is written for. Your own wording will differ:
| Deliverable | What it holds | Who it is written for |
|---|---|---|
| Executive summary | The recommendation, the value behind it, complexity, risks, constraints and next steps. | Leadership |
| Buy-in brief | A forwardable brief explaining why this initiative, why now, and what decision is being asked for. | The executive sponsor |
| Solution recommendation | Deployment model, architecture fit, technical assumptions and implementation considerations. | CTO, architecture and IT |
| Cost estimation | Token demand, model options, volume assumptions, year-one spend and the economic drivers underneath them. | CFO and finance |
| Implementation roadmap | A phased execution path from foundation through pilot, optimization and scale. | PMO and delivery lead |
| Viability assessment | Feasibility, data readiness, organizational capacity, timeline realism and a blocker analysis. | Risk and governance |
| Stakeholder alignment | Ownership, approvals, influence, adoption needs and an accountability map. | The change lead |
| Strategic add-ons | Governance, the ROI business case, a change plan, a training plan, the pilot charter and compliance depth. | Security, compliance, HR and finance |
Iternal publishes a full sample output set, readable before you book anyone's time. Read one rather than counting the list.
What the Tool Leaves to You
A blueprint scopes a project; it does not procure one. Iternal is direct about the edges: the output stops short of a full bill of materials, because the variability there is too wide, and it does not size cloud instances down to specific compute types. Some interview questions cannot be answered in the room and send the answerer off to research. Your compliance and technology experts still finish the plan, and its cost figures stay a range until someone quotes the work. Those edges define the last mile: teams that get value from a blueprint treat it as a scoped starting position they proofread and complete.
Answered elsewhere
- How a team finds and ranks its own candidate opportunities — see choosing your first use case.
- What a delivered engagement covers and how the phases run — see what an engagement covers.
- Starting when nobody in the building has done this before — see starting without an in-house AI team.
- Turning the numbers into a funded decision — see building the AI business case.
- For more information on where to begin — see where to start with AI.
FAQ
FAQ: The AI Blueprint Builder
The general use case exploration commonly runs 15 to 30 minutes; the deep dive on one use case commonly runs 45 to 90 minutes, and an hour or two is normal. Iternal describes a technical person who knows their own business finishing a deep dive in roughly 30 to 45 minutes, while a first-timer runs longer and a complex process longer still. Block the longer end.
Up to eight executive-ready documents, depending on the track you ran and how completely you answered: an executive summary, a forwardable buy-in brief, a solution recommendation, a cost estimation for finance, an implementation roadmap, a viability assessment, a stakeholder alignment map, and a block of strategic add-ons carrying governance, the ROI business case, a change and training plan, the pilot charter and compliance depth. Individual deliverables commonly run 10 to 20 pages, so ask for the current list rather than working from a count.
Yes. Iternal states the tool can be scoped by organization and by logged-in user, so an account can carry a directive to recommend only one hardware line, to stay inside a chosen catalog, to recommend cloud platforms only, or to skip cloud entirely. The default is agnostic: Iternal software is named only where it fits, and requirements outside the portfolio are scoped as custom development. Confirm which setting governs your instance.
Answers can be modified and a blueprint regenerated, which keeps the existing deliverables and adds newer ones alongside them, though editing one section currently replaces the rest of that section. Everything downloads and edits as a Word document, and exports as plain text, markdown, HTML and PDF. Ownership turns on editing: AI-generated text cannot be copyrighted until a person changes the wording, and once someone does, it can be, which is how partners re-paper a blueprint under their own brand.
No. Iternal describes a non-technical person running the interview by reading the questions off the screen, because the session needs business-case knowledge of the account rather than deep technical expertise. The real requirement sits elsewhere: the interview cannot be completed by someone who does not know the customer and how a large organization is structured. Put someone in the chair who knows the business.
Read the Output Before You Book the Interview
The fastest way to judge a guided-discovery tool is to read what it wrote for someone else. Open a sample deliverable set, find the sections your own program would live or die on, and decide whether an hour of interview is a fair trade.