Harness tightens up AI ROI spend with new tools
Software delivery company Harness produced two new products recently.
The directly-named AI DLC Insights and Cloud & AI Cost Management arrive with a promise of their ability to give software engineering teams real-time visibility into every penny spent on AI and whether that investment is producing measurable outcomes.
According to magical analyst house Gartner, worldwide AI software spending is expected to be $2.59 trillion in 2026, 47% more than last year’s spend.
According to Harness’s 2026 State of Engineering Excellence report, 94% of engineering leaders say the metrics that matter most are missing from their current measurement frameworks.
Unmeasured links
As spend continues to climb, the link between spend and business outcomes remains largely unmeasured.
“Every enterprise we talk to is asking the same question: we’re spending more on AI than ever, so why can’t we show what it’s doing for us? The first phase of AI adoption was about getting teams to use and understand the tools. The next phase is about proving the tools have a positive impact,” said Trevor Stuart, SVP and GM at Harness.
Stuart says that demonstrating ROI will be the defining challenge of enterprise AI in 2026 and that today, developers write nearly every line of new code with AI assistance.
The tools vary – Claude Code, Cursor, GitHub Copilot, Windsurf – but the pattern is universal.
The problem is that token spend has never been connected to outcomes i.e. what fraction of AI-generated code actually ships, how much spend is wasted on abandoned code or bloated prompts, and whether AI-assisted work is actually moving faster through review into production.
On-machine developer agent
Looking at AI DLC Insights, the product extends Harness Software Engineering Insights with a new on-machine developer agent that runs directly in the developer’s environment. The agent captures every AI-generated line of code, records token costs per model and tool, and maps that spend through the full delivery chain to the pull request, ticket, and deployment it produced.
The result is a complete picture of developer AI ROI: which tools teams actually use, where tokens are going to code that never ships, and whether AI-assisted work is producing faster, better software.
Key capabilities include: Unified AI coding adoption visibility: A single view of adoption across every major coding agent; Per-developer attribution: Token spend, sessions, and shipped code traced to the developer, team, and business unit; Wasted spend detection: Surfaces abandoned code, bloated prompts, expensive model choices, and missed cache hits; Coding-to-production impact: Tracks AI-generated code from prompt to production, using ship rate, PR cycle time, and DORA metrics correlated with incident data.
There is also benchmarking and governance, which compares team performance against org-wide baselines with role-based access control.
Live agents, new cost equation
Once an AI agent ships to production, a different cost equation takes over.
Every customer interaction, resolved ticket, and automated workflow triggers inference. In most organisations, spending is visible only at the invoice level, which tells us which line item is growing, but nothing about whether the growth is worth it.
“Cloud & AI Cost Management makes that determination possible,” said Stuart and team. “The product extends Harness Cloud Cost Management to cover every dollar of AI infrastructure spend, connecting directly to AI providers and production agents to capture spend at the individual request level and tie it to the agent, session, or workflow that triggered it.”
AI DLC Insights and Cloud & AI Cost Management are available in beta now.
