Skip to content
Alejandro García Startup Engineer_

AI process diagnosis

Let’s find where
AI improves
your results.

First I understand how your team works today. Then I show you, with numbers, which concrete options you have to work better.

~60 min · online · you get a free report

Sound familiar?_

  • Orders that arrive on WhatsApp and get entered twice: in a spreadsheet and in the system.
  • Customer data scattered across loose spreadsheets.
  • A key task that depends on a single person.
  • Errors and corrections nobody gets to measure.

You don’t need more technology: you need to look at the whole process before choosing one.

AI is the means. The goal is a better process.

The free diagnosis_

A meeting to understand how you work today and where AI can help.

Goal
Understand how you work today and where AI can help.
Length
About 60 minutes, mostly listening to you.
What you get
A free report with concrete improvement options.
Recording
The meeting is recorded so no detail is lost, and it is used only to build the report.

What we talk about_

First I listen, then I propose. Five topics, in this order.

  1. 01 · Scope

    Where do we focus?

    • Which area or process do you want to improve?
    • What takes your team the most time?
    • Where does work get stuck? Is anyone overloaded?
    • What would you like to stop doing by hand?
  2. 02 · Mapping

    Step by step, how it’s done today

    • Who does each step and how long it takes.
    • Through which channel: WhatsApp, email, spreadsheet or system.
    • How the information moves to the next step.
    • Screenshots, spreadsheets or manuals help: the report gets more precise.
  3. 03 · Numbers & risks

    How often it happens and what can fail

    • How many times per day or month, and how long each cycle takes.
    • How many errors or corrections there are.
    • What happens if the owner is away, and where information gets lost.
    • Whether there is sensitive data in chats or spreadsheets.
  4. 04 · Options

    We think together about how to improve it

    • Can it be automated?
    • Do you need a system, off-the-shelf or custom?
    • Can an AI agent read, decide and enter it for you?
    • Is training the team to use AI day to day enough?
  5. 05 · The goal

    What would success look like?

    • Where you are today and where you want to get.
    • Measured in time per cycle, errors, hours per week or customer response time.

What you get_

This is what the diagnosis report you take away looks like.

Diagnosis report

Wholesale distributor · Sales order entry process

Illustrative example

01 / Scope

From the moment the salesperson gets the order on WhatsApp until it is entered in the management system.

02 / Mapping

Every order is entered twice: in a spreadsheet and in the ERP.

  • 18minutes
  • 14steps
  • 32fields
  • 3systems

Risks found

  • Orders lost in the chat
  • Customer data in loose spreadsheets
  • Order entry depends on one person

03 / Options

  • A
    AI agentReads the WhatsApp order and pre-fills it in the ERP.
  • B
    Single formThe salesperson enters it once and the order arrives complete.
  • C
    AI trainingThe team uses AI to review and organize orders.

How we measure the improvement_

We compare before and after. That’s why the diagnosis measures how you work today: those numbers are the baseline that proves the improvement.

Option A · AI agent · Illustrative example

−72%

time per order

The agent reads the WhatsApp order and pre-fills it in the ERP. Admin only validates.

MetricBeforeAfterChange
Time per order 18 min 5 min −72%
Process steps 14 6 −8
Manual fields 32 4 −28
Systems touched 3 1 −2
Errors per 100 9 2 −78%

How I work_

Four phases, each with a clear deliverable. The diagnosis is the first one, and you decide at every step whether we move on.

  1. 01 · Diagnosis We start here

    We understand how you work today and propose options

    • We define the scope together: which process, which areas and who takes part.
    • I map the process with the people who run it every day.
    • A report with the current state (steps, times, manual tasks and risks) and options with cost, time and expected improvement.

    You choose the option.

  2. 02 · Design

    We turn the option into something you can see and validate

    • A system: a functional prototype with clickable screens.
    • An AI agent: the decision-and-action flow, with examples of how it responds.
    • An automation: the flow map with inputs, outputs and conditions.
    • Your users, decision-makers and tech team validate it. We don’t move on without everyone’s OK.

    Approved design, ready to build.

  3. 03 · Execution

    We implement, measure before and after, and train

    • Development following the validated design, integrated into how you work.
    • Real-world testing and comparison against the baseline.
    • Training for everyone in the process, not just leaders.

    New process running + documentation.

  4. 04 · Continuity

    Follow-up and improvements as the business changes

    • Follow-up with real users until the process is stable.
    • Adjustments as your systems or business change.
    • New improvement proposals as they come up.

    A stable process with an owner.

How I approach every project_

01

Understand before suggesting

First I listen and document how work gets done today. Proposals come after, not before.

02

AI is the means, not the goal

I use it where it adds value. Sometimes the best improvement is a simple automation or a process change.

03

See the solution before paying for it

A clickable prototype or a drawn flow: you validate it with your team before anything gets built.

04

Measure before and after

The initial assessment is the baseline. At the end we compare real times, steps and errors.

After the diagnosis_

  1. 1 We build the report If there are a few minutes left, we build it at the end of the meeting and go over it together.
  2. 2 I send it to you Right after the meeting, always: whether we went over it or not.
  3. 3 You choose and we move on Down the path that best fits your case.

Once you choose

Quote

If the solution is defined: fixed scope, timeline and cost.

Discovery

If a system needs to be designed: a fixed-price discovery and, at the end, the build quote.

Alejandro García

Who I am_

I’m Alejandro García, a Systems Engineer (UTN). For 19+ years I’ve designed and built software: I founded a software factory, co-founded a SaaS and was Product Owner at startups. Today I consult on processes and AI automation for companies in Argentina and Mexico.

  • 19+years building software
  • 15+MVPs for startups
  • AR · MXconsulting clients
See my CV and experience

FAQ_

Does the diagnosis cost anything?

No. The meeting and the report with the options are free. Only if you decide to move on do I send you a fixed quote.

Do I need to know about AI or technology?

No. You know how your business works; the technical side is on me. Everything is explained in time, steps and errors.

What if AI isn’t the answer?

I’ll tell you. Often a simple automation, an existing tool or training the team is the better choice.

Why is the meeting recorded?

So no detail is lost. The recording is used only to build the report.

What should I have at hand?

Screenshots of your systems, the spreadsheets you use and, if there is one, a process manual (no sensitive data). If nothing is ready, we share screens during the meeting.

Do you work with companies outside Argentina?

Yes. Everything can be done online; I currently work with clients in Argentina and Mexico.

Book your diagnosis_

Write to me and we’ll find a time for the meeting.