Opsera today added an artificial intelligence (AI) reasoning agent to its platform that is specifically trained to surface in real time how DevOps teams are deriving value from investments in AI coding tools.
At the same time, Opsera is adding to its namesake platform for managing and automating DevOps workflows an ability to integrate with the Model Context Protocol (MCP) server developed by GitHub.
Finally, the company is also making an on-premises edition of its platform, dubbed Insights in a Box, available to IT teams that are not able or inclined to rely on a software-as-a-service (SaaS) platform for managing DevOps workflows. This self‑hosted edition of an existing Opsera Insights service that pulls data across the ecosystem, normalizes it, and surfaces key performance indicators (KPIs) that align with DevOps Research and Assessment (DORA) metrics.
Opsera CEO Kumar Chivukula said the Hummingbird AI Reasoning Agent makes it possible to determine which developers are generating code using AI tools that make it into a production environment. It tracks usage and adoption of AI coding tools, impact, value and productivity metrics to also surface optimization recommendations, he added.
That capability is critical because the more developers employ AI coding tools, costs start to rise as more tokens are consumed. The intelligence surfaced by Hummingbird AI Reasoning Agent can then be used to, for example, allocate more tokens to developers that are generating higher quality code, noted Chivukula.
Hummingbird AI agent understands context as it reasons across systems and tools to surface recommendations for improving productivity in natural language, said Chivukula. DevOps teams can chat with their own data to, for example, ask “why” something happened, “what” to do next, or “how” to improve performance. Those queries are then shared with large language models (LLMs) developed by Open AI and Anthropic, with the best answer then being shared back to the DevOps engineer, he added.
While AI coding tools are already widely adopted, generating more code doesn’t equate to more applications being deployed in a production environment. Many organizations are now generating more code but the developers reviewing that code don’t always understand how it was constructed. Often, that code can not only contain vulnerabilities, it is often overly verbose. As the volume of that code increases, the potential amount of technical debt being taken on increases as more inefficient code is added to an application environment.
In effect, Opsera is making a case for a DevOps platform that now uses AI to identify coding issues being created by other instances of AI.
It’s not likely that the AI coding genie is going back in the bottle any time soon, so the challenge now is determining how best to manage DevOps workflows when the amount of code of varying quality being generated has exponentially increased. Ultimately, the only way to achieve that goal is to rely on other instances of AI to provide greater visibility into software development lifecycles (SDLCs) where the number of AI coding tools being employed to write code is only going to continue to increase.
