Harness has introduced two new merchandise designed to assist organizations achieve higher perception into the prices and advantages of AI usage. The options handle a rising drawback inside IT departments: AI is being deployed extra regularly, however it’s typically unclear what these investments truly yield.
In line with a Harness survey of engineering managers, 94 % haven’t included value metrics in how they consider software program growth, studies SD Times. Consequently, there’s a lack of visibility into whether or not AI instruments contribute to larger productiveness or, quite the opposite, primarily generate extra prices.
The primary new product, AI DLC Insights, focuses on software program growth. The answer tracks what number of tokens builders devour through AI code assistants and hyperlinks that knowledge to concrete outcomes, resembling resolved bugs, pull requests, or new performance. This enables organizations, for instance, to find out how a lot AI capability is required to carry out a particular process and whether or not that’s more cost effective than conventional growth work.
Harness observes that builders are producing increasingly more code utilizing AI, whereas a few of it finally by no means makes it into manufacturing. The platform goals to disclose the place AI investments truly result in usable software program and the place cash is wasted on unused code or inefficient processes.
The software program helps a number of in style growth instruments, together with GitHub Copilot, Cursor, Claude Code, and Windsurf. Moreover, prices and outcomes are damaged down by particular person builders, groups, and enterprise items. The platform additionally flags conditions the place token consumption will increase with out a corresponding rise in software program manufacturing.
Perception into operational AI prices
Along with growth environments, Harness additionally goals to offer management over the operational prices of AI providers. To that finish, the corporate is introducing Cloud & AI Value Administration.
This resolution consolidates AI spending from varied distributors right into a single overview. This enables organizations to trace prices for providers from suppliers resembling OpenAI, Anthropic, AWS Bedrock, and Google Vertex AI. In line with Harness, this overview is usually lacking, that means that finance departments obtain invoices however battle to evaluate which functions add worth.
The platform hyperlinks bills to particular AI brokers, workflows, groups, and enterprise items. This makes it clear which functions are working effectively and which of them are literally incurring excessive prices with out demonstrable outcomes.
The answer additionally consists of options for funds administration and price monitoring. Organizations can set limits for groups or tasks and obtain alerts when spending rises unexpectedly. That is supposed to stop a brand new AI software from quietly resulting in a sharply rising invoice.
In line with Trevor Stuart, senior vp at Harness, the most important drawback proper now isn’t the extent of AI spending itself, however the lack of perception into its return on funding. With the brand new merchandise, the corporate goals to assist organizations handle AI investments in the identical means they handle cloud prices: primarily based on measurable outcomes reasonably than assumptions.