What is our primary use case?
My main use case for Mistral AI is the chatbot functionality, which I find really great and very fast compared to ChatGPT and Claude.
I use Mistral AI chatbots in my work to ask questions, particularly simple questions.
When I say simple questions, I don't perform Excel work or financial data analysis in Mistral AI; I use Claude for that and prefer Claude. However, for anything related to regulations or accounting rules, I prefer Mistral AI because it is more straightforward and direct.
Mistral AI can bring different value to the market by really focusing on being efficient and being straightforward. I feel a lot of people are frustrated with AI chatbots that dance around the point and don't give real answers sometimes in complex issues, so I think Mistral AI is doing a good job in this aspect.
How has it helped my organization?
Mistral AI helps me gain a lot of efficiency and productivity. As soon as I have a question, I get the answer in real time, which is really beneficial. Before, we were going to consulting firms about tough questions and paying a lot of money, but now I can just ask Mistral AI and it will give me a good answer.
What is most valuable?
In my opinion, the best features Mistral AI offers are that it is based in Europe and on European servers, so the data is not really shared.
This has an influence on my organization because we are originally a French company, so it's important to keep our data close rather than sending it to US big tech firms.
Regarding Mistral AI's capabilities, I think the governance and security are good, especially compared to the other actors out there. Since it's based in Europe and in France, I feel there is better governance and security.
I think the accuracy and reliability of Mistral AI's output are good, and I have never had to complain about this.
What needs improvement?
Mistral AI has a lot of work to do on image generation and Excel functionality. Building dashboards and Excel work for professional use is something I see as needing improvement since it is currently seen more as a chatbot and is less advanced than Claude in this aspect.
Ideally, what we would love to see is artifacts similar to what we have in Claude and being able to create our own AI agent with Mistral AI because we want everything in one place. We don't like the fact that sometimes we have to change from one chatbot to another, and if we had one place where we could do everything, it would be the best.
For how long have I used the solution?
I have been using Mistral AI for a little bit more than a year.
What was our ROI?
We don't have that data yet in terms of measurable return on investment, but in terms of time, it's pretty obvious. Instead of doing a one-hour call to understand an issue, now in five minutes I have my answer. Also, in terms of email generation, it saves a lot—I would say more than 10 hours per week.
What other advice do I have?
Mistral AI is a good tool, but there is still development needed, and it has a lot more to do in terms of integrations, dashboards, data analysis, and the visualization of the analysis.
I purchased Mistral AI through the AWS Marketplace.
I would advise others looking into using Mistral AI to really adapt the tool to their industry and make sure that it fits their issues and their day-to-day problems.
I rate Mistral AI a six out of ten.
Which deployment model are you using for this solution?
Private Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
