I'm working in an AI solutions company with no data analytics focus. We are developing solutions with Microsoft tools.
AI projects have gained customer trust and integrate smoothly with existing workflows
What is our primary use case?
What is most valuable?
Azure AI Bot Service is a valuable feature we utilize. Microsoft's security and governance capabilities make it easier for us to sell to our customers. Market penetration and interoperability with the Microsoft ecosystem are significant benefits. Customer trust is another important factor, as customers tend to be less demanding with Microsoft compared to competitors like Alteryx. When customers have selected the technology and trust Microsoft as a partner, they tend to be less aggressive when it comes to problems, issues, and bugs.
What needs improvement?
The solution is not the most mature product. The choice was made because Microsoft holds approximately 95% of the market in Southeast Europe, and all customers request Microsoft products. There are actual bugs in the way it works that cause breakdowns. Consistency and reliability are significant issues, as it doesn't have the expected reliability in the way it returns information and data. The language models are behind, particularly with non-English languages such as Greek, Romanian, and Bulgarian. There is a real issue with non-English language support that needs to be addressed. We see different consistency across regions and times, with performance dropping significantly after 3:00 or 4:00 PM when US users access the system. More stability, reliability, consistency, and better reasoning in non-English languages would be beneficial.
For how long have I used the solution?
I have been using this solution for three years.
What do I think about the stability of the solution?
There are actual bugs in the way it works that cause breakdowns. Consistency and reliability are significant issues, as it doesn't have the expected reliability in the way it returns information and data.
What do I think about the scalability of the solution?
Consistency across regions and times is problematic, as performance drops significantly after 3:00 or 4:00 PM when US users access the system. More stability, reliability, and consistency are needed.
Which solution did I use previously and why did I switch?
I previously worked for an Alteryx provider as an integrator but switched jobs and am no longer using Alteryx.
Which other solutions did I evaluate?
At the moment, we are not exploring other options because the market demands Microsoft products. In my region, there is no discussion about alternatives as customers want Microsoft because they have Office 365 and desire interoperability.
What other advice do I have?
Azure AI is the primary focus. I would rate the customization capabilities as eight out of ten, as you can do custom things. The maturity level is not very strong based on what we have observed. We intend to explore additional features in the future. Overall, I would rate this solution eight out of ten.
Which deployment model are you using for this solution?
Public Cloud
