Problems

Not all complex problems need the same solution.

The first step is to recognize what kind of situation actually exists, which uncertainties prevent progress, and what level of intervention makes sense.

We do not know how to approach it

What is usually happening

There is a real need
There is no clear technical or organizational path
Several possible alternatives appear
It is not yet clear whether the solution requires AI, software, process changes, or another approach
There are constraints or dependencies that are not yet well understood

Questions worth resolving

What problem are we really trying to solve?

What constraints shape the solution?

Which part of the problem requires investigation?

Which alternatives are reasonable?

What evidence do we need before committing to a direction?

What may emerge

Diagnosis
Conceptual model
Solution alternatives
Architecture
Roadmap
Prototype
MVP
A grounded decision not to continue

We have already tried and it does not work

What is usually happening

There is a partial solution
The result does not respond to the original problem
The solution only works for one specific case
It does not scale
It is difficult to maintain
A PoC cannot evolve
The architecture limits growth

Questions worth resolving

What was the initial solution trying to solve?

Which part does work?

Where is the cause of the blockage?

Is it a problem of concept, data, architecture, integration, operation, or process?

Is it better to evolve what exists or rethink it?

What may emerge

Critical review
New conceptual model
Revised architecture
Simplification
Process redesign
New validation strategy
Rethought MVP

We need to build a new capability

What is usually happening

The organization wants to be able to do something it cannot do today
There is not yet an architecture, process, or governance model
The capability needs to combine people, procedures, and technology
An isolated project is not enough

Questions worth resolving

What should this capability exist for?

What decisions should it enable?

What processes does it need?

What knowledge and profiles are required?

Which part should be built, bought, or integrated?

How can it be validated before industrializing it?

What may emerge

Strategy
Governance
Operating model
Processes
Architecture
Platform
MVP
Capability roadmap

We do not know where the best opportunities are

What is usually happening

There is interest in AI, automation, or data
Many ideas are generated
Prioritization is missing
It is not clear where real value exists
The organization does not know what to explore first

Questions worth resolving

Which problems or decisions have the greatest impact?

Where is there enough information or data?

Which opportunities are technically feasible?

What does it cost to experiment?

What should be validated before investing?

What may emerge

Opportunity map
Prioritization criteria
Portfolio of initiatives
Roadmaps
Experiments
Selective PoCs
Decisions not to invest

The goal is not to generate a generic list of artificial intelligence initiatives. It is to identify where there is a relevant problem, a reasonable hypothesis, and a proportionate way to test whether there is value.