Assessment One workflow · one clear decision

Before you automate it, decide if it belongs.

The AI Workflow Readiness Assessment examines one recurring work problem. You leave knowing what is actually broken, what AI can and cannot do, what must remain human-controlled, and whether a small pilot is worth proving.

This is not a prompt workshop or an automation build. It is a disciplined decision before you invest in either.

What is broken

Map the recurring work problem, the current workaround, and where context or ownership is failing.
Identify the information the workflow can actually trust.

Where AI fits

Define whether AI should prepare, compare, surface, classify, or draft.
Specify the human review point, decision boundary, and fallback.

What happens next

Receive a clear recommendation to repair first, stop, or test a defined pilot.
Leave with success criteria that prove improvement instead of activity.

MethodVisible before you buy

Six questions that keep an AI workflow owned.

Work problemWhat keeps getting stuck, repeated, or repaired?

OwnerWho is accountable for the result and exceptions?

Source truthWhat information can the workflow safely rely on?

AI roleWhat should AI prepare, compare, surface, or draft?

Human boundaryWhat must a person decide, verify, approve, or own?

ProofWhat would show the work became better, not just busier?

Fixed-scope assessment USD 4,500 One workflow. No artificial scarcity.

One workflow, one accountable owner, up to two source systems, and representative examples. The work takes five to ten business days from a confirmed kickoff. Payment is 50 percent at kickoff and 50 percent on delivery. Before kickoff, we send a short written agreement covering scope, payment, data handling, and working terms at least two business days before the scheduled call. You may request changes before signing. If we decline the work before kickoff, there is no charge. If you stop after kickoff, we deliver the completed work and do not charge the remaining balance. If scope changes, we pause and agree written options before doing more work.

You receive

A 3 to 5 page decision memo with the work problem, owner, source truth, and recommendation.
An annotated workflow map with the AI role, human review boundary, and fallback.
A repair-first checklist or pilot test card with success, risk, and stop criteria.
A 45-minute readout and one written clarification round. Submit clarification questions within ten business days of delivery; the response clarifies the work already done and does not add new research. You own the deliverables and may use them with any team or vendor.

Your team's time

One 60-minute kickoff, one or two 45-minute working sessions, and one 45-minute readout.
Ten to fifteen representative examples. Live system access is not required unless we agree it is useful.
A source system means one distinct system of record, such as a CRM or shared drive. Email or chat can be supporting context unless it is the record. If more than two are essential, we provide written options to narrow scope or revise price. No extra charge happens without your agreement.

Data and no-fit terms

An NDA is available on request. DXDEV can provide a standard form or work from yours. Raise this before submitting so we can confirm terms fit the timeline. Do not submit credentials, raw customer data, or regulated records through the form. Read the plain-language data handling note.
Client material is not used to train models. We do not send it to third-party AI tools without written approval.
If the workflow is not ready, the paid output is a usable repair-first or stop recommendation, not a pressured pilot.

Founder-led. The founder leads the assessment and the written decision. The free AI at Work library and Build Logs show the operating method behind the offer.

Independent fit check

Ask your own AI whether this is useful.

Before you contact us, paste this into any AI assistant. It should help you test the fit, name what is still unclear, and tell you if a lighter next step makes more sense.

I am considering a DXDEV AI Workflow Readiness Assessment for one recurring work problem. Act as a skeptical operations advisor, not a salesperson.
  
  Based on the details I add below, tell me:
  1. Is this work problem specific enough for an assessment?
  2. What may need repair before AI is involved?
  3. What should AI do, and what should stay human-controlled?
  4. What source of truth, owner, review step, and fallback would be needed?
  5. What measurable improvement would make this worth changing?
  6. Would a $4,500 fixed-scope assessment be useful here, or is there a simpler next step?
  
  Do not recommend the assessment just because I mentioned it. State uncertainty clearly and give me three questions to ask DXDEV.
  
  My work problem:
  Owner:
  Information sources:
  Current workaround:
  Risk if wrong:
  What better looks like:

Fit check Six short answers

Bring us one workflow that keeps getting stuck.

We begin with the problem, not a tool. These answers tell us whether the work has an owner, usable context, a review path, and a meaningful way to measure improvement. We reply within one business day.

Your name

Your email

Company

Your role

1. What recurring work keeps getting stuck, repeated, or repaired?

2. Who owns the result today?

3. Where does the information come from? (optional, helpful)

Name the documents, systems, conversations, or records people currently rely on.

4. What is the current workaround? (optional, helpful)

5. What could go wrong if the system is wrong? (optional, helpful)

6. What measurable improvement would make this worth changing?

Is AI already involved? (optional) No, we are deciding whether it belongs here. Yes, but the work is not yet reliable or well-owned. We are not sure what is already being used.

Anything else we should know? (optional)

Intake submissions are transmitted through Web3Forms and reviewed by DXDEV. Web3Forms privacy terms govern handling during transmission. DXDEV retains the submission only as long as needed to assess and respond, normally no more than 90 days. We do not share it or use it to train models. Read the plain-language data handling note. Please do not include credentials, raw customer data, or regulated records.