Decision support from Locollabs AB
How to choose an AI consultant
Start with the work that needs to change, who is responsible and how the result will be tested. These seven questions help you compare proposals before introducing AI.
Locollabs provides AI advisory and implementation services. This is our approach to assessing an engagement, not an independent ranking of suppliers.
1. Which task needs to improve?
Describe who does the work, which information they use and what goes wrong or takes too long. A clear problem lets you compare an AI solution with a simpler process change.
2. What can our existing tools do?
Ask the consultant to assess what you already pay for. Custom development needs a reason, such as a remaining need or a requirement your current environment cannot meet.
3. What counts as an accepted delivery?
Ask for concrete test cases, named users responsible for testing and written acceptance. A working demonstration does not establish that the solution works in everyday use.
4. How will we measure value?
Establish a baseline and compare similar tasks. Include preparation, review, corrections and your own team’s time. Time freed up creates capacity; it does not automatically become a cash saving.
5. Which data and actions are permitted?
Clarify access, storage, model providers and decision-making authority. If personal data is processed, roles and responsibilities need to be assessed for the actual use case.
6. Who maintains the solution after handover?
Ask for documentation, export options, operational responsibilities and exit terms. Understand the cost of maintenance and what happens when a model or integration changes.
7. What is included in the price, and can we decide not to proceed?
Separate an introductory call from a paid assessment, implementation and ongoing support. A sound assessment may recommend waiting. That needs to be a possible conclusion from the start.
Useful reference material
The Swedish Authority for Privacy Protection’s guidance offers a starting point for questions about AI and personal data. NIST’s voluntary framework can help structure risk management. Both need to be applied alongside an assessment of your own solution.
Want to apply these questions to your situation? Read how Locollabs approaches AI adoption or arrange an introductory call with David Mancilla.
Book an introductory call