Right now, the main use cases for Gemini involve mostly doing experimentation on agentic AI use cases, context understanding, making sure that the context is clearly understood, and then in RAG-based applications.
The role that we use Gemini for is mostly doing the heavy lifting on different agentic AI work that we do. For example, context understanding, making sure that it generates the right result, interpreting it, summarizing it, all those things.
I have seen benefits in terms of enhancing data flow or management within my organization. In terms of Gemini, it is mostly being used to do the agentic AI work.
What makes Gemini stand out is the multimodal feature that they introduced. It was one of the first solutions where users can generate images, which is actually pretty cool.
The main benefits that Gemini brings to the table include definitely speeding things up significantly. It is also introducing many new use cases that we were not able to work on earlier. The automation part of it is fantastic.
The code generation capabilities in Gemini could be improved. When comparing it against Anthropic, they have better code generation capability.
There are specific features that should be included in Gemini in the future. Code generation exists but sometimes debugging is not as good. For Google-related technology, it works very well, but for other code, it sometimes hallucinates.
I have been working with Gemini since it was released, approximately a year ago. I have been using it extensively for different activities through APIs as well as directly through the web interface.
My advice or recommendation to other users who are looking into Gemini is that there are many options available, but from its core capabilities, especially multimodal capabilities, and from the agentic AI framework, Gemini is doing very well from that perspective. On a scale of 1-10, I rate Gemini a 9.
