How to Choose AI Service Provider and What Questions to Ask

How to Choose AI Service Provider in 2020

Most likely, you no longer need to read about the potential of artificial intelligence for business. Therefore, let’s start this article with a little experiment. Right now say “Ok Google, which developers are the best AI companies?” And see how many companies position themselves as top AI companies and strive to convince you to turn to them for the best solution. Obviously, it is necessary to seriously approach the issue of choosing AI service provider, so in this article, we have put together a checklist of questions that must be asked at the negotiation stage to filter out unreliable companies from the list of potential vendors.

 

You Have a Problem, They Should Have a Solution

If you want to introduce artificial intelligence because your competitors have already done this, it may be partly correct, but there is one thing. An AI-based system should solve a specific problem of your business, for example, saving HR managers from routine work through the implementation of a smart HR system.

What is more, you must clearly understand how you will measure the effectiveness of this solution, and the AI service provider, based on this, should not provide you with a choice, but an individual solution tailored to your needs.

 

Is the Company Immersed in Your Business? 

In continuation of the previous paragraph, in order to provide you with the most suitable solution, AI service provider should immerse yourself in your business and your problem as much as possible. Until they feel your need from beginning to end, they will not be able to create anything worthy for you. You, as a business owner, will always be able to assess the depth of such a dive and immediately weed out those companies whose first initiative was superficial.

 

What AI Solutions Do They Use in Their Business Processes? 

Here is another good point to check the reliability and seriousness of the company. It is obvious that anyone who develops AI solutions like SPD Group should use them to improve the efficiency of their own business. Otherwise, it is like a shoemaker who s always ill shod. Feel free to ask a direct question about which smart systems are used, what tasks they solve and what results they bring. Serious and honest companies usually do not hide this information, of course, omitting technical and commercial details, which is also normal.

 

What Is Their Real-Life Experience? 

If you have already clarified which AI systems are used inside the company, now is the time to find out about what the company created for other types of business. The best cases are usually kept in the portfolio in the public domain, but no one forbids you to ask additional questions and find out if the company has developed something similar to what you need. If they had such projects, then the project manager will easily tell you about their experience and will show the result, if, for example, the developed application is available for everyone and is not a secret tool for a certain company or state body.

 

Get to Know the AI ​​Development Team

Of course, this is your legal right to find out who will work on your project, or even get to know each member of the team in person. The most ideal option is when the company has its own in-house team – everyone is in the same place, at the same time and working on the same project. But keep in mind that this most reliable approach means a higher development cost due to staff maintenance.

At the same time, many companies successfully work in the format of remote teams and are no less productive than in-house ones. Therefore, evaluate this criterion comprehensively. If you understand that the company is suitable for you in all other respects, then perhaps the remote form of interaction is not something to find fault with. The main thing is that communication should be established. But more on that in the next paragraph.

https://pixabay.com/photos/computer-laptop-technology-office-3368242/

How Communication Is Configured? 

Let’s make a reservation right away – well-functioning communication is important for both in-house and remote teams in the development of artificial intelligence systems. And in each of the options, its extremes happen. For example, an in-house team may be completely unorganized, while a remote team may be able to resolve current issues in a split second.

Therefore, in each case, ask a direct question – how do team members communicate with each other, and how will they communicate with you as a customer? It is normal practice to coordinate reporting forms, software for monitoring the work of a remote team, as well as tools and time for communication between you and the team at the start of the project.

 

What Technologies Do They Offer and What Are the Advantages/Disadvantages for You

The technologies used are one of the key factors that will form the final price of the project. But this is not even the most important. The most important thing is that the technologies that will form the basis of the product will determine its performance and effectiveness for the business. Therefore, there are a couple of important points.

If you yourself are a little versed in the technical intricacies of creating artificial intelligence systems, then, of course, you can make your suggestions at the initial stage of negotiations. However, paradoxical as it may be, this is an alarming sign if your offers are unambiguously accepted. Remember we talked about real experiences?

So, the one who has real experience always has his own vision and ideas on how to create this or that AI project. And is not afraid to voice them, in a soft form, of course. In addition, when all your offers are peremptorily accepted, this may be a signal that, in fact, the company is still quite new and they simply agree to all the conditions so that you would not go to their more famous competitors. This leads to the conclusion – the more personal expertise from the team you will see at the start, the more likely that you will get a quality product at the end.

 

What Is the Cost

The issue of price is always very sensitive in itself. And first of all, you as a business owner should understand how you will take this amount out of profit or reserves without harming the business, but rather, in order to invest in AI.

In addition, most likely it will be impossible to estimate the cost of the project with an accuracy of a penny, as it is impossible to estimate the necessary time with an accuracy of several minutes. Therefore, do not be alarmed if you hear something like this. This is a fairly normal approach, and now we will explain why.

When it comes to practice, your project will be transformed and this is normal. You will want to introduce some more improvements and this is also normal. Therefore, the best thing you can hear is a detailed explanation of what bottlenecks and complexities might appear in your idea, with the caveat that this may take extra time and require additional money.

If the AI ​​service provider warns you in advance of potential difficulties, this is a good sign. This is simultaneous evidence of experience, serious intentions, a reasonable approach multiplied by the desire to discuss all possible details before the start, to save you from surprises.

But in any case, you have the right to know the minimum price, which is quite realistic to calculate.

 

Let’s Check It Once More

So, let’s look again at the checklist on how to choose an AI service provider in 2020.

  1. Relate your problem to the solution that you are offered and ask to show quantitative evidence of the value the company is going to deliver.
  2. Evaluate whether the company delves into your business processes
  3. Feel free to ask questions about already implemented projects and AI systems that are used inside the company.
  4. Meet the team members in person and determine how effective the communication between them is.
  5. Get ready to receive counter offers
  6. Do not be afraid to hear the preliminary price.

 

Bio: Helen Kovalenko is IT Project Manager at SPD Group. She is working in a Data Science Team over the NLP, Computer Vision, and Fraud Detection solutions, and gladly shares her researches and developments in her blog posts.