00Latitud IA14°00′

The model was never the problem.

Latitud is the system we use to locate an organisation: which decision deserves to improve, which capability is holding it back today, and how far it needs to go. Latitud IA is its first instrument, the one that measures the artificial intelligence question.

The starting point
The wrong question

Which model should we buy?

The right question

Which decision do we want to improve, and what has to change for the improvement to last?

Por qué importa

An organisation can own the best model on the market and change nothing. It happens every day, and the cause sits in the operation: the model arrives into a structure that was never built to receive it.

Holding a great deal of data and trusting none of it. Running documented processes governed by informal exceptions. Installing a capable copilot outside the workflow. Training the team while leaving every incentive untouched. Automating a decision nobody can answer for when it fails. Launching pilots nobody measured.

Latitud IA exists to find those fractures before they turn into wasted investment. It names the organisation's position with evidence, points to the capability holding it back, and states how far it has to go. Not one level less, not one more.

The system

Four coordinates fix the position.

Like any coordinate system, it locates rather than grades. No organisation is good or bad at artificial intelligence. It sits at a point, and we can measure that point.

01

North

Which decision has to improve, who answers for the outcome, and what its economic logic is.

Without North: attractive pilots that move nothing.

02

Evidence

Whether the data the decision demands exists, holds up to scrutiny, and can live inside the real systems of the operation.

Without Evidence: plausible answers built on false data, or demos that never reach production.

03

Decision

How the operation produces the outcome today, which rules and exceptions govern it, and who answers when the system gets it wrong.

Without Decision: automating the mess and scaling the error at machine speed.

04

Adoption

Whether the organisation has the capacity, the incentives and the will to fold the system into how it works.

Without Adoption: the right technology, installed, and nobody uses it.

The rule

Your position is set by the weakest link.

Almost every maturity model averages its dimensions, and that is where they stop being useful: a comfortable average hides the one capability that will sink the project.

In Latitud IA the level a decision can support is the lowest of its critical coordinates. An organisation with impeccable North and Evidence stays out of production if nobody can answer who stops the system when it fails. Weighting sorts investment. Someone has to clear the block.

One capability at zero cancels the result, however excellent the model.

The scale

Six levels, and no obligation to reach the last one.

  • L0

    Opacity

    The organisation does not know how its own outcomes are produced. The knowledge lives in people's heads, not in systems.

  • L1

    Exploration

    The organisation understands itself and experiments, but each area tries on its own and nothing connects.

  • L2

    Foundation

    Critical processes defined, baseline metrics in place and owners with names. There is ground for a supervised pilot.

  • L3

    Operation

    Intelligence lives inside the workflow, with active controls and results measured against a baseline.

  • L4

    Scale

    A governed portfolio, reusable components and learning that travels from one area to another.

  • L5

    Adaptation

    The organisation learns continuously and redesigns its decisions with evidence. Autonomy proportional to risk, never absolute.

Every decision has a level that is enough for it.

Some operations work well at L2, and pushing them to L4 burns capital to decorate an org chart. Traditional consulting sells the whole staircase because the whole staircase is its business model. We state the level the decision demands and count stopping there as a success.

The instrument

Move the four coordinates. The lowest is your position.

Mark where your organisation stands today on each one. The lowest one rules, not the average: whatever the other three have to spare sits waiting on it.

The instrument stores nothing and sends nothing. The maths happens in your browser.

Drag each coordinate to the level you recognise today.

1 of 4
01NorthL0L1L2L3L4L5OpacityExplorationFoundationOperationScaleAdaptationLimiting coordinate
02EvidenceL0L1L2L3L4L5OpacityExplorationFoundationOperationScaleAdaptationLimiting coordinate
03DecisionL0L1L2L3L4L5OpacityExplorationFoundationOperationScaleAdaptationLimiting coordinate
04AdoptionL0L1L2L3L4L5OpacityExplorationFoundationOperationScaleAdaptationLimiting coordinate
Your positionYour position: L0-A. Limiting coordinate: Evidence.OpacityExplorationFoundationOperationScaleAdaptationThe organisation does not know how its own outcomes are produced. The knowledge lives in people's heads, not in systems.Limiting coordinateNorthEvidenceDecisionAdoption
North is your constraint: whatever you build will run without anyone able to say which outcome improved.Evidence is your constraint: any system on top will answer with confidence about data nobody verified.Decision is your constraint: automating today freezes into code the informal exceptions holding your operation together.Adoption is your constraint: you will pay for a correct system that your team works around to carry on as before.
Stranded capabilityNorth
Confidence of the evidence
We saw the data, the process or the control working.There is backing, incomplete or limited to part of the operation.Someone in the organisation asserts it and it is recorded with that weight.Still without backing. The level stays suspended.

The coordinates you moved are stated evidence: grade C, at best. The diagnosis verifies it on the ground, and a verified L2 carries more weight than a stated L2.

The verdict

Five possible exits. One of them is not doing it.

The diagnosis forces a single thing: deciding with evidence. Five possible paths come out of it.

A

Strategic clarity

There is no problem, owner or clear metric yet. The decision gets defined first.

B

Evidence foundation

Data is the block. We prepare the minimum trustworthy set the case needs, not an endless platform.

C

Process redesign

Exceptions govern the work. Automating now would freeze the mess into code.

D

Vertical AI

The case clears the readiness required. We build the copilot, the agent or the decision system.

E

Do not implement

Value is low, risk disproportionate or evidence insufficient. The organisation's capital and attention are protected.

Recommending against implementation is the cleanest proof of independence a firm can give. A consultancy that lives on selling implementations cannot issue that verdict. We can, and we have put it in writing inside our method.

The deliverable

A diagnosis that reads without us in the room.

The report stands on its own, with no consultant beside it explaining. It is a document for acting.

  • 01The current position and the confidence of the evidence behind it.
  • 02The map of the four coordinates, with the constraint marked.
  • 03The level the decision demands, and why not more.
  • 04The case portfolio sorted into Now, Prepare, Explore and Discard.
  • 05The risk class, the controls and who answers for each one.
  • 06The 90 to 180 day route, with owners and investment decisions.
  • 07And the next decision the executive committee has to approve.

Every finding states its own evidence.

A level observed directly weighs differently from one stated in an interview, and the report marks it with a letter: verified, partial, stated or unproven. The report keeps what we saw apart from what we were told. We publish only the precision the evidence holds.

Specification of the artificial intelligence maturity diagnosis.Method sheet
Latitud IA · Cohorte

You bring a real decision. You leave with the route.

Six to eight organisations go through the Latitud IA diagnosis in parallel, each on a real decision from its own business. They leave with their position measured and their route written.

It is our consulting method, run with the organisation in the room instead of handed over in a report. Four modules, one per coordinate. Each module closes with a decision on the case the organisation brought.

The format exists for a practical reason: the cohort spreads the cost and puts eight organisations through the same uncomfortable questions at the same time.

0

Organisations per edition, at most. The value is in working one real case per company, and that does not scale to an auditorium.

0

People per organisation: whoever decides and whoever operates. One alone is not admissible, and that rule is not negotiable.

0

Modules, one per coordinate. Each closes with a decision on your own case, not with an exam.

0 days

The route every organisation leaves with, including owners, gates and the next investment decision.

You come in two at a time.

Each organisation sends the executive sponsor and the process owner. Signing and operating are different crafts, and transformation falls into the gap left between them. The four modules exist to close it in the room. Anyone sending a single person stays out of the Edition.

Los cuatro módulos
INorth

Which decision deserves to improve

Each organisation arrives with a decision from its operation and leaves with it bounded: the outcome that has to change, who answers for it, the baseline and the economic logic that justifies it.

Closes by deciding whether the case deserves to continue.

IIEvidence

Whether the data holds the weight

Fitness, ownership, access, quality, semantics, lineage and protection of the data the decision demands. And whether whatever the team builds can live inside the real operation or will die as a demo.

Closes knowing which evidence is missing and who will produce it.

IIIDecision

How the decision works today, and who answers

The real flow end to end, with its rules and its informal exceptions. What intelligence will assist, what it will automate and what it should never touch. Who supervises, who overrides and who stops it.

Closes with the risk class of the case defined.

IVAdoption

Whether people will use it

The pain the user recognises, the incentives that sustain or sabotage the new behaviour, the support and the telemetry that will show whether anyone uses it. The coordinate that buries the most projects and receives the least planning.

Closes with the verdict: pilot, prepare or stop.

What each organisation takes away

  • 01Its one page diagnosis: position, confidence of the evidence and decision.
  • 02Its limiting coordinate identified, which is almost never the one expected.
  • 03Its 90 day route, with owners and the committee's next decision.
  • 04The vocabulary to hold the conversation internally, without us.

Organisations share the method, the vocabulary and the questions. Each one keeps its data, its numbers and its diagnosis, and signs that commitment before the first session.

What the Cohort is not

A cohort diagnosis locates and prioritises. Without observing the operation on the ground, much of the findings will rest on what the organisation states, and the report will mark it with its letter. Anyone who needs evidence verified on the ground needs a mandate. We say it on day one.

Edition I

Edition I runs in Quito, in Spanish. If your organisation wants one of the places, or if you want to propose yours as host, write to us and we will talk about the case you would bring.

City
Quito
Language
Spanish
Places
8 organisations
Format
4 in person modules
Application

Two names, or the application does not move.

The place belongs to the organisation, and the organisation is two people: whoever sponsors the decision and whoever operates the process. Both fields are required because the rule is real.

01Executive sponsor
02Process owner
Direct questions

What clients ask about the method.

01

What is Latitud IA?

It is Equator Consulting Group's proprietary system for diagnosing artificial intelligence maturity and deciding the transformation route. It assesses four coordinates (North, Evidence, Decision and Adoption) against a concrete business decision, places the organisation on a six level scale and delivers a 90 to 180 day route with the next investment decision.

02

How does it differ from other artificial intelligence maturity models?

In three ways. Your position is set by the weakest link, because that is the one that sinks the project. The diagnosis targets a concrete decision: the same organisation can be ready in finance and not yet in human resources. And it states the level that decision demands, which is almost never the top of the scale.

03

Can the diagnosis conclude that we should not implement artificial intelligence?

Yes, and it is one of the five formal outcomes of the method. When value is low, risk disproportionate or evidence insufficient, the recommendation is not to implement. It protects the organisation's capital and it is the proof that the diagnosis was not designed to sell the next project.

04

What is Latitud IA Cohorte and who can take part?

It is the diagnosis run as a cohort: six to eight organisations go through it in parallel, each on a real decision from its own operation. Each organisation takes part with two people, the executive sponsor and the process owner. A single person is not admissible, because the distance between whoever decides and whoever operates is what makes transformation fail.

05

How long does the diagnosis take and what does it cost?

We fix scope, duration and price before starting, and they do not move along the way. Be wary of anyone who quotes a figure without knowing the decision you want to improve. The first conversation costs nothing and is usually enough to tell whether the work makes sense.

06

Does Latitud IA work for public institutions?

Yes, and the Decision coordinate carries more weight there than on any other ground: traceability and accountability are the condition of legitimacy for a public decision, long before they are an audit requirement. The method is the same; what changes is the governance demanded of it.

If your organisation invested in artificial intelligence and sees no change, the diagnosis starts by knowing where it stands.