One week. You'll know what's actually working, what's theater, and what to ship next. On your data — not ours.
of companies creating substantial value from AI
have seen no real return on AI investments
of teams stuck in pilot purgatory
Sources: BCG Build for the Future 2025 · Typeface / Content Marketing Institute 2025 · McKinsey State of AI 2025
You're probably not where you thought you'd be with AI. Neither is almost anyone else. This diagnostic identifies where the gap is for your company specifically — and what to do about it.
Not a technology audit. We're not evaluating your stack or recommending tools.
Not an IT or infrastructure project. Data architecture isn't the starting point.
Not a software-selection exercise. Anyone leading with vendor evaluation is selling something.
Not a training programme. We don't run workshops on prompts.
Not a report that goes on a shelf. You leave with a working demo on your own data.
Five artifacts. Clarity → Plan → Proof → Decision → Handover.
The honest picture of where the company stands. Business map of how money flows. Team-by-team AI state snapshot — who's using what, how often, who isn't, why not, what's paid for and unused, what's working quietly, what's theater. Gap analysis of what should be happening at this size/stage/industry and isn't.
3–5 prioritized opportunities, each with problem, agent sketch, data/integrations needed, effort estimate, ROI hypothesis with a High/Medium/Low confidence band. A 90-day sequenced execution plan (what ships in weeks 4, 8, 12). An architecture blueprint so the agents form a system, not 12 disconnected tools. Build engagement pricing so you can say yes on the spot.
A crude prototype of the top-priority opportunity, running on your sample data, live and clickable by end of week. Not production-grade — proof-grade. This is the deliverable that separates Agentropic Impact Diagnostic from every consultancy's equivalent. A consultancy writes a report. We ship a prototype.
Ninety minutes. Working session, not a handoff. We walk the leadership team through the Impact Report and Build Plan, pressure-test priorities together, edit live, land on a decision before the call ends. You leave owning the plan, not receiving it.
Everything in writing — final report and build plan edited per session, demo access link, short walkthrough video for the CEO to show the board, pricing for next Build engagements.
A four-beat arc. Every engagement should end with you able to say all four things out loud about your own company.
Interviews, tooling audit, business map. By end of day 3, we can describe your company's AI reality better than most of your team can.
Opportunities scored on Impact × Feasibility × Fit. 3–5 picked, sequenced, costed. Architecture drawn.
Build the top-priority demo on your sample data. 1.5–2 days of build, time-capped. Live and clickable.
90-minute Live Executive Session. Pressure-test, edit, decide what ships next — before the call ends.
Effort split: ~3.5 days investigation + ~1.5 days demo build + ~1 day synthesis + session. Team: 2 people minimum, always Ashish + one other on delivery.
Heads tell you policy. Operators tell you reality. We talk to both, every time.
Always. Sets the top-3 priorities the Build Plan has to answer.
Where distinct from a function head. Tech reality and integration constraints.
Support, finance, ops, product, whoever we're mapping. They tell you policy.
They tell you reality. The single most valuable finding in most engagements comes from watching someone actually do the work, not hearing someone describe it.
$5K floor. Effort-based above that. Parameters: company size, functions in scope, output depth, industry complexity.
| Scope | Interviews | Duration | Price |
|---|---|---|---|
| Small (1–3 functions) | ~8 | 1 week | $5–12K |
| Mid (4–6 functions) | ~12 | 1.5–2 weeks | $15–30K |
| Full org (7+ functions) | 18+ | 2–3 weeks | $30–60K |
50% at start · 50% on delivery of working demo + report. No refunds. No free pilots.
If the diagnostic leads to a build, you choose who operates it.
Your infrastructure. Your team runs the agent.
$8–18K per agent · ~$2K/yr maintenance
We ship the code to your repo on day 1. Perpetual unlimited license. Buyout at 2–3× build cost if you want the source outright.
We run the agents on your behalf — hosting, monitoring, prompts and models evolved as they improve.
$1–3K/month per agent
Priced by scale and traffic. Cancel anytime with 30 days' notice.
Or take the Build Plan and ship it without us. The plan is yours. Larger multi-function transformation priced separately.
The diagnostic may or may not lead to a Build engagement with us. That decision is entirely yours — the value of the diagnostic is the clarity it creates, not the relationship it implies.
Tell us the company, the rough scope, and when you want to start. We'll send the NDA, the intake list, and a Monday kickoff slot.