AI Development Lifecycle installs agent software on developers’ machines to track adoption, sessions, and code created across coding agents. It identifies whether licensed AI coding tools are used or whether shadow AI tools are being used instead. It traces tokens consumed and code shipped to production by developer, agent, repository, team, and business unit, surfacing wasted tokens from abandoned code, bloated prompts, expensive model choices, and cache misses. DevOps teams can measure AI-generated code using pull request cycle time, incidents, vulnerabilities, and DORA metrics. Cloud & AI Cost Management extends cost tracking for AI infrastructure and flags spending spikes before invoices using an existing detection engine.
"An AI Development Lifecycle (DLC) tool installs agent software on a developer's machine to track adoption, sessions, and the code created across every coding agent, while a Cloud & AI Cost Management tool has been extended to track spending on AI infrastructure."
"For example, the AI DLC provides insights into whether developers are using the AI coding tools that organizations have licensed or are using a set of shadow AI coding tools instead. Additionally, the number of tokens consumed and code shipped to production can be traced to the developer, agent, repository, team, and business unit. Tokens burned on abandoned code, bloated prompts, expensive model choices, and cache misses are all surfaced automatically."
"DevOps teams can track AI-generated code based on pull request (PR) cycle time, number of incidents and vulnerabilities created or any of the DevOps Research and Assessment (DORA) metrics. Finally, spikes in spending are also flagged before they hit the invoice using the same Harness detection engine that many DevOps teams already use to monitor cloud costs."
"While there is little doubt that more code than ever is being generated, most DevOps teams have little insight into how much cost is actually being incurred or how much of that code is making its way into a production environment. Too many organizations have simply been focused on encouraging developers to adopt these tools, which often leads to them being used to generate as much code as possible."
Read at DevOps.com
Unable to calculate read time
Collection
[
|
...
]