Finding the right AI vendor must be tough!

AI is here. It works. Since it has many uses in all parts of your business, in all industries, choosing an AI vendor must be tough. After all, there are so many!

We broadly see 3 types that help. One is who has the technology to build and manage models that power software applications. The second is one who builds software applications that use such models. The third is who offers services by using several technologies and applications. Be clear about what you are dealing with as there is also the fourth type that tricks you in confusing services for a product. Or worse, a fifth type that just masquerades legacy software as Artificial Intelligence to get attention!

So how do you go about separating fiction from reality?

Here are the 4 questions to ask.

#1. Can I see?

Insist to not use a presentation to talk about the greatness of AI but rather show the real thing. How do build models? How to deploy. See.

#2. Can it do?

On seeing a demo, ask to show results using your own data. Check the time and effort it took. And how it was done. This gives you a realistic estimate of the skills that are required.

#3. Can I do?

AI is not what you buy once but rather what you continuously build upon. Check if you can repeat the vendor’s work to do it yourself. Would you have to go back to the vendor for changes?

#4. The roadmap?

Technologies made a few years back are already old. Does the roadmap support your future? Be specific. Ask what and when.

Hiring a vendor is one thing. But getting a partner is another.

Is the vendor committed to your success? Does it share a similar vision? It is easy to check that by the effort they are putting into your projects and in developing similar businesses in your niche. It is a good sign if key people are available to meet you and work on your problems.

We find that cooperation shouldn’t just be just a seller-buyer relationship, but you have to share the common goal of making each other successful.

Start modeling now


Attend a training