Unbabel - The AI Translation Platform

My main use case for Unbabel is language translation. A specific example of how my team uses Unbabel for language translation is in the situation of translating data in different languages. We used to put that information into Unbabel and then retrieve the translations, mainly used for Portuguese and German. It helped us in that situation, assisting us with translation and providing the best information on the use case.

I have more to add about how I use Unbabel, noting that the overall interface was slow, and many of the words were not translated properly. Additionally, I used to check against Google Translate to double-check if the translation was proper or not. Overall it was a good experience, and you have to train the AI to get a better result.

In my experience, the best features Unbabel offers and the parts that stand out to me the most are that the translation part was easy to navigate and easy to use. I can pinpoint the information which I was looking for, making it easy to draw attention to the Portuguese and the German translation, which were better.

I would add that everything was good regarding the features. The translation part was easy and takes a little bit of time, but it was good. The translation part was up to 99% accurate, and you can easily find the information you are looking for, so it's good.

Unbabel has positively impacted our organization by addressing the cases where the number of cases was a lot more because of the language barrier. We were not able to solve the cases so fast, but Unbabel came, and we put the language into it, and the translation was happening. It was good to have the translation ready on the call, as you just click on it and get the part done easily and quickly.

I can give an example of how much time I was able to save; earlier, we were able to do ten or fifteen cases in an hour, but then it increased to twenty. The five cases per hour increase was because of the translation since it was easy to get the translated things done and much handy to use.

Unbabel can be improved by being a little bit faster; I think the user interface can be a little bit more interactive. The slowness of Unbabel is there, and you need to wait a little bit; it hangs sometimes. If the platform can be a little bit more user-friendly and fast, it would be better.

I wish for additions to the platform that could make it even better, such as more languages being added, which can enhance the overall experience.

To address any other improvements Unbabel needs, I feel everything is overall fine, but if slowness and accuracy can be improved a little bit more, it will be more helpful.

I have been working here for three and a half years in my current field.

Unbabel is stable; I have not experienced any downtime or issues with reliability. It is always up and available for use anytime; the servers and everything work properly, making it reliable.

Unbabel's performance when handling large volumes or spikes in translation requests does slow down; when you put in large amounts of data, it slows down, and it sometimes takes longer than usual time. You have to wait, and it keeps getting slow if you put large data sets together, making it troublesome. You must ensure you break it into bunches to use it effectively.

Unbabel's scalability is limited; it does not handle large amounts of data efficiently. If you put too much data into it, it gets hung, so scalability is an area that needs to be tested and worked on.

The customer support for Unbabel is acceptable; we shared the continuous usage reports with the client, and they sometimes had to reach out to support. However, we were not in direct contact with the support team.

We were not using a different solution before Unbabel; we used general solutions such as Google Translate and some local websites to get translation results. We tried to switch to one professional platform to gather all results in one place.

My experience with the initial setup process with Unbabel was challenging; things were not working properly, and access was confusing. If you have to click here and there and the process was not working, it sometimes gets hung. In the beginning, it was trouble, but slowly you get used to it. The setup was hard initially.

I would describe Unbabel's integration with our existing tools or workflows as easy to connect; it was straightforward. As I mentioned, in the beginning, it was trouble, but once it was done with access and troubleshooting was complete, then it was easy to access, integrate, and use.

I manage user access and permissions within Unbabel flexibly according to our organization's needs and requirements. We were able to get the access we required, and the organization was able to provide it, making user requirements and permissions easy to obtain as well.

Unbabel's reporting and analytics are indeed helpful; any kind of analytics provided are beneficial in general. They help the team understand the results and what is happening. Team managers were able to share the insights, which were helpful for us as well.

I would describe the training and onboarding process for new users in Unbabel as initially not being enough. As I mentioned, it was a new product for us, so the training provided was insufficient. The team slowly gathered the SOPs and other information that helped us get the desired results. In the beginning, we faced unclearness in the SOPs, but as time went on, it improved.

I have seen a return on investment from using Unbabel; after using it, the time saved was significant. Before, we had to go to different websites to get translations, but after using Unbabel, we were able to get all translations in one place. This saved time and increased efficiency since we could handle five more cases in an hour, which was also helpful for productivity.

I did not evaluate other options before choosing Unbabel because I was working for an organization, and that was the only project they had. Therefore, we did not explore other options.

Regarding Unbabel's AI capabilities, I think its governance and security comply with the laws, including GDPR, and its security is good. As I understand, the AI compatibility is in its starting phase; slowly, it will increase, but right now it needs development.

I need to double-check the output always; when I was using it, I used to double-check the translated output I received against Google Translate. I cannot rely one hundred percent because some of the words I knew were translated wrong. The AI has to increase its accuracy; I find it to be about ninety percent reliable, with ten percent requiring double-checking to ensure correctness.

We ensure compliance with data privacy and regulatory requirements while using Unbabel by making sure sensitive information is not copied and pasted, so right-clicking is disabled. This prevents copying any sensitive information, which is against GDPR. We also inform clients that their data could be used to help the organization get better results or for translation. We communicated that we would conduct testing to achieve results.

I was part of the organization using Unbabel, so I do not have details on pricing, setup costs, and licensing for Unbabel; I only worked on the Unbabel project.

My advice for others looking into using Unbabel is to avoid inputting all the information at once since the tool is still in development and growing. I would suggest not relying on it one hundred percent initially; just use it, and once you are familiar with how it works, it will be easier to use.

My overall review rating for Unbabel is seven out of ten.