Where AI agents are showing real IT savings
LaunchDarkly saved $1.17M by deploying coding and IT support agents — but one expert warns: most orgs are inflating savings by ignoring the hidden cost of human oversight.
Why it matters
IT agents are delivering measurable ROI in narrow, well-defined domains (tier-1 support, cloud cost optimization, coding tasks), but real savings are 30–40% lower than pilots suggest because organizations under-count ticket reopenings, human verification labor, and workforce transition costs.
The key facts
8 to knowLaunchDarkly: $1M cost avoidance from coding agents + $120k from internal tool builds instead of software purchase + $50k annualized from tier-1 IT support
West Monroe Partners: 40% reduction in MSP costs; 2,700 hours/year operational time savings from IT/HR support agent in Teams/Slack
KamiwazaAI: first round of cloud optimization agents cut spending by 70% immediately
Jeet Pattanaik (Glokal AI) skepticism: Tier-1 example with 40% closure rate and 15% reopens yields ~60% of promised savings after verification labor is counted (€36k actual vs €60k claimed)
Pattanaik: ticket reopening rates and human checking costs are rarely tracked in business cases
Use case sweet spot: high-volume, well-documented, low-cost-to-undo work (tier-1 passwords, access resets); tier-2 and permission changes remain high-risk
Hidden cost identified: tier-1 automation eliminates junior learning path, reducing pipeline to senior engineers
Multiple CIOs still hunting for ROI; metrics vary across organizations
Go to the source
CIOcio.com
Publisher excerpt: Many IT leaders still struggle to find ROI when deploying AI agents at scale, but a few IT sweet spots are emerging as use cases that can provide CIOs tangible cost relief. Rhonda Baldwin, CIO at LaunchDarkly, says coding agents and service desk agents are helping the SaaS provider cut IT costs.…