The bill always comes due. And for enterprises that spent the past two years racing to deploy artificial intelligence, that reckoning arrived faster than many chief financial officers expected.
PointFive, a three-year-old cloud efficiency startup founded by the team behind cybersecurity firm IntSights, closed a $60 million Series B round on June 8—a vote of confidence from Accel and a roster of returning investors that the market for taming runaway infrastructure costs has room to run. The New York- and Tel Aviv-based company has now pulled in $96 million since launching in 2023, a trajectory that mirrors the urgency with which enterprises are suddenly scrutinizing their cloud and AI expenditures.
The catalyst? A reversal that caught many by surprise. After years of steady improvement, cloud waste ticked up to 29% in 2026, according to Flexera's annual State of the Cloud report. It's the first increase in several years. The culprit isn't hard to identify: AI workloads, with their voracious appetite for GPU clusters and large language models, have outpaced the tools designed to keep spending in check.
"Companies are discovering their bills are climbing faster than their dashboards can track," one enterprise architect at a Fortune 500 retailer told me recently, speaking on background because they weren't authorized to discuss vendor relationships. It's a sentiment that PointFive's founders—Alon Arvatz, Gal Ben-David, and Amir Hozez—heard often enough to build an entire business around.
When the Previous Exit Opens Doors
The trio's credibility precedes them. IntSights, their cybersecurity venture, sold to Rapid7 in 2021 for $335 million. That kind of exit buys you more than just capital for the next act—it buys legitimacy with enterprise buyers who remember what you built before.
It shows in the customer list. Nubank, the Brazilian digital bank. NICE, the enterprise software giant. E.ON, the European energy conglomerate. Koch Industries' Infor division came aboard after what Salesforce Ventures described—perhaps generously—as "portfolio development efforts" in a November 2025 blog post. Translation: introductions were made, relationships were leveraged.
Accel, which led the Series B alongside participation from Index Ventures, Salesforce Ventures, Entrée Capital, Vesey Ventures, Sheva Ventures, and Perpetual Growth, frames the opportunity as "$1 trillion in optimization" across cloud and AI spending. TD Cowen's analysts project that GenAI public cloud expenditures alone will quadruple between 2025 and 2028. Goldman Sachs pegs Big Tech's AI capital expenditures at $700 billion this year, climbing toward $1 trillion in 2027.
Those upstream investments cascade downstream to the enterprises running their own infrastructure, often with limited visibility into where the money actually goes.
More Than FinOps
PointFive doesn't want to be lumped in with traditional FinOps platforms—the spreadsheet-adjacent tools that help finance teams track cloud spending but often lack teeth when it comes to actually fixing anything. The company bills itself as a "Cloud & AI Efficiency Engine," covering more than 500 optimization types across AWS, Azure, Google Cloud, and Kubernetes environments. As of March, that expanded to data platforms: Snowflake, Databricks, BigQuery.
In February, the startup launched DeepWaste AI, a module aimed squarely at production AI spend. It targets LLMs, GPU infrastructure, and the data platforms that feed them. The pitch is agentless, read-only integrations that promise setup in under five minutes—no agents to install, no security reviews that drag on for months.
From there, the platform offers what the company calls "agentic remediation": one-click scripts and infrastructure-as-code patches that route directly to engineering teams via Jira tickets, ServiceNow requests, Slack messages, or GitHub pull requests. It's an attempt to bridge the gap between identifying waste and actually eliminating it, a distance that has long plagued the FinOps category.
Accel's investment note claims customers see "average cost savings of 30% and ROI up to 1400%," though those figures come from the investor's portfolio data, not independent audits. (Venture capital firms have been known to round up.)
A Crowded Field Gets More Crowded
The broader market for cloud cost optimization remains dense with competitors. Apptio Cloudability, CloudZero, Vantage, Flexera, CloudHealth—all jostling for FinOps budgets. Kion just added Anthropic token spend management on June 4, a signal that the competitive landscape is evolving fast toward AI-specific cost controls.
PointFive's wager is that enterprises don't want point solutions for each layer of their infrastructure. They want a unified platform that spans traditional cloud resources, data platforms, and AI workloads in a single pane of glass. Whether that integrated approach resonates with FinOps teams already managing multiple vendors—and whether those teams have appetite for yet another dashboard—remains an open question.
The company has grown to more than 100 employees across offices in New York, Tel Aviv, and London. It joined AWS's ISV Accelerate program in July 2024 and maintains a presence on Azure Marketplace, signaling a go-to-market strategy tightly coupled to the hyperscaler ecosystems. That makes sense: if you're selling cloud efficiency, you might as well sell it where the cloud lives.
What the Money Buys
PointFive plans to use the Series B proceeds for product development and expanded sales and marketing efforts—the same language you'll find in nine out of ten funding announcements. More interesting, perhaps, is what the round signals about investor appetite for infrastructure efficiency at a moment when AI spending is spiraling.
Previous rounds include a $20 million Series A in November 2024 led by Salesforce Ventures and a $16 million seed in May 2023 led by Index Ventures. The escalating check sizes suggest investors believe the market is expanding, not consolidating.
Whether PointFive can translate that belief into sustainable differentiation in a crowded field—and whether its founders' second act proves as lucrative as their first—is the $96 million question. For now, at least, they've convinced a room full of venture capitalists that the bill is only getting bigger, and someone needs to help enterprises pay it.
