Ler original em português

Published

AI Consulting or Software House? Deciding What to Do Is Not the Same as Building

By Pedro Sonda, dooop

An AI consulting firm helps decide what to do with AI, in what order, and under which rules; a software house builds and integrates the chosen solution. If the company does not yet know which problem to tackle, the expected value, or which risks to accept, the first step is consulting. If the use case is already validated and needs to become a system, the next step is the software house. Many companies need both, in sequence.

Updated in October 2026.

Transparency note: dooop is an AI consulting firm, not a software house.

The Difference in a Table

AI ConsultingSoftware House
Question answeredWhat to do, why, in what order, under which rules?How to build, integrate, and maintain?
Typical deliverablesDiagnosis, prioritized portfolio, roadmap, governance, trainingSoftware, integrations, APIs, maintenance
TeamStrategy, business, processes, risk, educationEngineering, data, architecture, product
How success is measuredBetter decisions, prioritized value, adoptionWorking system, on time and on scope
Risk when used at the wrong timePlans without executionBuilding the wrong solution well

Why the Order Matters

Gartner predicted that at least 30% of generative AI projects would be abandoned after proof of concept by the end of 2025, citing poor data quality, inadequate risk controls, rising costs, and unclear business value (Gartner, 2024). Of these four reasons, only one is essentially technical. The other three are decisions that should come before building.

BCG found that only 22% of companies progressed beyond proof of concept with AI (BCG, 2024).

Interpretation: Building more proofs of concept without selection criteria does not solve this bottleneck.

dooop Framework: The 3 Question Test

Answer before hiring:

  1. Do we know which business problem we want to solve and how much it is worth?
  2. Do we know what data, risks, and rules are involved in this problem?
  3. Do we know who will use the solution and how their work will change?
  • Three "yes" answers: look for a software house (with governance oversight).
  • Any "no": start with consulting; the "no" indicates which work to hire: discovery (1), governance (2), or adoption and training (3).

The Sequence That Reduces Waste

Decide → Validate → Build → Adopt

StageWho LeadsDeliverable
DecideConsultingProblem, expected value, criteria, and priority
ValidateConsulting and technical teamExperiment or proof of concept with success criteria
BuildSoftware house or internal teamIntegrated solution
AdoptConsulting and business areasPeople using it, adjusted processes, indicators

The adoption stage is often overlooked in development contracts. Without it, finished systems remain underutilized.

What If the Same Provider Offers Both?

It can work well, provided that:

  • The strategic recommendation is not tied to selling development;
  • The company has clear criteria to accept or reject the build recommendation;
  • The proposal separates price and deliverables for each stage.

What to Do Now

To learn about dooop’s approach, see dooop.com.br/consultoria.

Frequently Asked Questions

Can a software house do AI strategy? Some do. Verify if the team has experience in business decision-making and if the recommendation is independent of what they sell.

Does AI consulting develop systems? It depends on the provider. Consultancies focused on strategy, governance, and training usually do not develop but can conduct proofs of concept and support hiring builders.

Also Read

About dooop

dooop is a consulting firm based in Caxias do Sul (RS) that prepares leaders and organizations to transform artificial intelligence into business capability through strategy, governance, and training.

Share your context, constraints and the decision your organization needs to assess.

We will use your details to deliver this content and contact you about related topics.