StitcherAI weaves threads to assess whether AI ROI is dressed for success

FOCUS first

First, it has a semantic engine, built on FOCUS (the open billing standard adopted by AWS, Azure, and Google, co-created by StitcherAI’s founder), which unifies cloud, AI, SaaS and vendor data, and models cost in the business terms boards actually ask about: products, customer segments, margins, unit-economics KPIs, forecasts and cost controls. 

That intelligence flows continuously into data lakes, BI platforms, JIRA, Slack, and ERPs an organisation already runs on.

Omnipresent reasoning engine 

Second, an omnipresent reasoning engine of specialised agents steers organisation-specific financial context into agentic workflows and AI platforms, like Claude, Cursor, and Codex, as IT dollars are committed. 

That shift, from managing cost after the invoice to making business context aware decisions, transforms IT Finance from reactive to proactive, and cuts the time to understand business impact by months. 

Engineers and agents commit spend at a velocity no finance team or tool was built to track. An agentic workflow can indeterministically route to different tools, models and vendors, each with its own pricing and token math, so unit cost varies from request to request. Existing IT Finance tools depend on engineers remembering to tag, finance teams continuing to police IT spend, and stakeholders remembering to check dashboards.

Human scale, agent scale

At human scale that produced friction – at agent scale, it produces nothing at all.

Enterprises need a platform that understands their specific IT costs, complexities, and financial goals and can seamlessly drive intelligent decisions.

”I witnessed this problem firsthand while leading global IT Finance at Citi,” said Udam Dewaraja, founder and CEO of StitcherAI. “Humans and agents didn’t check the dashboards of existing tools when committing spend and often didn’t have access to all the relevant data. With StitcherAI, we reimagined how organisations manage modern IT spend using AI to automate low-impact decisions while providing business-aligned financial intelligence for humans to make higher-order IT investment decisions.”

IBM research suggests that just 25% of AI initiatives have delivered expected ROI over the last few years, and only 16% have scaled enterprise-wide. Engineers and autonomous agents in Claude Code, Cursor, and Codex increasingly make architecture and infrastructure decisions in seconds. 

At the same time, the IT Finance infrastructure governing them was built for weekly review. This creates a pattern where AI and IT spend gets committed with no consideration of ROI. Finance teams are then forced to firefight when the invoices arrive. 

“CIOs and CFOs are no longer asking, ‘how do we manage cloud and AI costs?’ They’re asking, ‘which AI investments are producing returns? How do we scale those and quickly kill the ones that aren’t?’” said Tim Crawford, CIO Strategic Advisor at AVOA and a member of The Wall Street Journal Technology Council. 

Crawford says that what CIOs and CFOs need is a system that puts their enterprise’s financial context into the workflow at the point of decision, for humans and agents. 

Unintended execution paths

The situation worsens as agentic workflows drift onto unintended execution paths or hallucinate in ways the original design never anticipated. Fortune 100 enterprises are spending 18 to 24 months and millions in labour costs to build in-house solutions to wrangle IT spend, but that’s too slow and expensive for most enterprises now facing rising AI and IT investments. 

FOCUS – FinOps Open Cost & Usage Specification is an open specification that normalizes billing datasets across AI, cloud, SaaS, data center, and other technology vendors to reduce complexity for FinOps Practitioners.