Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
Generations of developers have worked with command line interfaces both to build and deploy applications.
A challenge with that approach is that it requires knowledge and manual effort, taking time and skill, which are at a premium in the modern world. It’s a challenge that Amit Eyal Govrin saw time and again while he was working at Amazon, where he managed the devops partnerships for the cloud giant. Shaked Askayo faced a similar challenge, as he was working as a devops leader at fintech startup, BlueVine, and he needed a system that could replicate his skill set to help the organization scale up its devops efforts.
Askayo and Govrin got together in early 2022 and came up with the idea to build out a way to solve the problem they both had experienced. That idea became Kubiya, which is launching out of stealth today with $6 million in seed funding.
“Kubiya is an advanced virtual assistant for devops,” Govrin told VentureBeat. “Self-service devops is what we’re enabling.”
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
How conversational AI enables faster devops workflows
The basic idea behind all conversational AI technologies is that a user or a consumer can have a conversation with an artificial intelligence (AI) powered system that will provide answers, or help to execute a task.
Conversational AI is already widely used for bots and automated chat tools that respond to queries on all types of different public websites and services. Customer experience (CX) platforms are among the leaders at embracing conversational AI and it is also being used to help businesses stay organized with technologies such as Xembly. According to a report, conversational AI can also be a boon to supporting mental health treatment.
Kubiya is now taking conversational AI in a different direction, as a way to help developers with domain-specific knowledge as well as workflow. Govrin explained that Kubiya can be embedded into the tools that developers are using to help provide answers to questions they might have. More importantly from his perspective, though, is that Kubiya is looking to provide a self-serve platform that executes devops tasks.
“We’re allowing people to have full-length conversations that are converted into operational workflows,” Govrin said.
Artificial intelligence (AI) is no stranger to the world of devops, though it has often been used for different purposes than what Kubiya is doing. For example, there are AI-assisted development tools, including GitHub Copilot, which provide code suggestions for developers to build applications.
Govrin said that GitHub Copilot is all about code completion, in contrast to what Kubiya is doing, which is the next level and helping with operational completion, getting applications up and running in production environments. While there isn’t a direct integration between Kubiya and GitHub Copilot today, Govrin said it’s likely there will be some form of integration in the near future.
Conversational AI techniques that enable self-serve devops
Devops is not a single tool or a single operation. Rather, modern devops involves a complex configuration of different tools and services that organizations use to build, test, deploy and manage applications.
Govrin said that Kubiya integrates with commonly used tools today and is growing its capabilities on a daily basis with a bidirectional feedback mechanism. As such, when devops professionals encounter situations that Kubiya can’t handle, there is a mechanism to provide a suggested workflow operation that can then be added to the platform.
“We’re allowing our end users to be active participants to train the system,” Govrin said.
Kubiya has also developed an approach that Govrin referred to as multi-organization entity recognition. Govrin explained that the multi-organization entity recognition approach allows a user of Kubiya to provide as much or as little context as they want and the system understands what to abstract and what to ask to further clarify what context is missing.
For example, Kubiya can be used to help a devops professional to start a new cloud instance. The conversational AI system will understand the basic query and then also ask other relevant questions to help determine the intent of the query in order to properly configure the right type of cloud instance, with the necessary access and security controls.
“You can be as specific or as generic as you need,” Govrin said. “The system will recognize the context and abstract the context that it requires in order to get the rest of the information needed to execute a request.”
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.