What is our primary use case?
We were looking for a SRE assistant with Kubiya.ai that could help us with daily routine tasks by automating them or functioning as an AI agent that could perform actions for us. For example, if we received a customer ticket, we wanted to automate that process.
One specific example of a task that Kubiya.ai helps automate for our team is addressing ticket fatigue. We were removing the middleman work for our DevOps engineers. Instead of manually servicing repetitive tickets, such as restarting pods or granting access to particular resources, Kubiya.ai automated this through chat. Our developers gained self-service capabilities without needing to learn Terraform or CLI commands, and our DevOps team members could focus on architecture rather than support tickets.
What is most valuable?
Kubiya.ai functions as a DevOps or SRE assistant for us. It is not merely a chatbot or an LLM interface or exactly ChatGPT, but rather an action-oriented platform. Most AI bots simply chat, whereas Kubiya.ai is a complete agentic platform built to execute code. It connects directly to our Kubernetes, AWS, GitHub, and Jira and other tools to perform end-to-end workflows.
A particularly good example is our ability to spin up a dev environment for payment services. Kubiya.ai triggers the Terraform script, waits for completion, and pastes the URL back in Slack. This end-to-end workflow execution is exceptionally valuable. Additionally, Kubiya.ai is security-oriented. A common concern in AI for DevOps is that a bot could accidentally delete a production database. Kubiya.ai solved this with strict role-based access control and human-in-loop features. If a request appears risky, it can ping a manager for approval before executing.
We saved considerable time regarding productivity with Kubiya.ai. We did not need to hire as many resources for support tickets, and the process was quite smooth.
What needs improvement?
Kubiya.ai's billing structure is somewhat complex. The usage-based pricing charges per function call, which can lead to unpredictable bills if adoption spikes unexpectedly. Enterprise pricing is quite high. The setup complexity could also be improved. There are trust and hallucination risks, though these generally depend on how we use the AI.
For how long have I used the solution?
We have been using Kubiya.ai for the past year.
What do I think about the stability of the solution?
Kubiya.ai is stable. We have not encountered any issues so far, though we may have more insight in the future.
What do I think about the scalability of the solution?
We have not seen any scalability issues with Kubiya.ai. We do not have extensive data in this area, so we are currently satisfied.
How are customer service and support?
We have not reached out to customer support for Kubiya.ai. However, we can share feedback once we have experience with that service.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
We were using Kubernetes GPT as a previous solution. However, we needed something very quick and an enterprise solution, which is why we chose Kubiya.ai.
How was the initial setup?
My experience with pricing, setup cost, and licensing has been smooth. Currently, we are in the initial phase, so I cannot comment extensively on this area. We are still validating those metrics and will provide additional information once we have gathered more data.
What other advice do I have?
Kubiya.ai is best suited for large engineering teams that suffer from numerous repetitive support tickets. Teams that use Slack, such as ours, have benefited significantly from Kubiya.ai without requiring developers to learn new tools, and we can manage everything directly from Slack.
I would rate Kubiya.ai nine out of ten because it has completely revolutionized the way we operate in Kubernetes. I chose nine out of ten instead of a perfect ten because if you are a small startup or individual, the cost and setup overhead will likely outweigh the time saved. Additionally, you need a zero-trust environment where no automated system is allowed to touch production infrastructure without manual human execution. These factors depend on your specific use cases. In my personal opinion, the cost factor is something to consider heavily.
Others looking into using Kubiya.ai should try it. It is truly a good enterprise solution, and the way it is designed to execute end-to-end agentic AI workflows is exceptional, particularly if you need a SRE assistant. My overall review rating for Kubiya.ai is nine out of ten.
Which deployment model are you using for this solution?
Public Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
