$1 Trillion of AI and Cloud Spend will Need to be Optimized: Our Investment in PointFive
AI is transforming enterprises, but also creating new crises of governance, visibility and - urgently - cost. As usage increases, CIOs are finding they have no visibility of how AI tools are being used and CFOs are waking up to spiralling token costs they didn’t budget for. Global spend was over $350 billion in 2025 and is expected to surpass $1 trillion by 2030.* This is on top of ever-increasing cloud spend, which creates waste at the deepest levels of a system.
This lack of oversight over how AI is deployed is becoming a serious operational issue that enterprises need to address. In most cases, they don’t just need to know what they’re spending on AI - they need to optimize how their teams use it, eliminate waste, and monitor for compliance violations. The FinOps tools that enterprises use to monitor cloud usage aren’t set up for the level of visibility that this new era needs.
PointFive has developed a landmark new solution: a new cloud and AI optimization engine that tracks AI and cloud usage with unprecedented detail. Their thesis is that this infrastructure needs to be optimized by the engineers who are working day to day on it. PointFive helps engineers decide between performance vs. costs, and allows team leaders to see exactly when and how tokens are deployed. The result is real dollar savings made and costly mistakes avoided - customers already report average cost saving of 30% and ROI of up to 1400%.
We've known Alon Arvatz, Gal Ben David, and Amir Hozez from their days as the founders of IntSights, a cyber threat intelligence platform that was acquired by Rapid7 for $350 million. At PointFive, they have brought a security-style research approach to cloud and AI infrastructure management. Their in-house research team studies cloud and AI services the way a security team studies novel vulnerabilities. PointFive has built a catalogue of more than 500 inefficiency types and is adding 12+ new optimizations each week.
On the cloud side, these optimizations include uncovering misconfigured resources, over-provisioned compute and Kubernetes clusters, orphaned workloads, and inefficient database and storage configurations across all the major cloud providers, as well as Snowflake & Databricks. On the AI side, in addition to visibility of AI usage and cost across coding models, PointFive optimizes usage by condensing prompts before bills are generated and tagging compliance issues as they happen.
By showing teams what they are wasting and why, PointFive surfaces cloud and AI inefficiencies that conventional tooling misses entirely. Its combination of proprietary efficiencies and easy remediation has resonated with large enterprises such as E.ON, Infor, Nubank, and NICE. These enterprises are turning to PointFive to achieve the oversight and automation necessary to manage today’s AI and cloud infrastructure.
Today, we are thrilled to lead a $60 million round in PointFive alongside existing investors Salesforce Ventures and Entrée Capital. Alon, Gal, and Amir have worked together for over a decade, and their complementary skill sets, mutual trust and rapid execution speed make them uniquely placed to help enterprises manage the growing complexity of AI and cloud infrastructure, which we believe will be one of the largest challenges of the AI era.
*Survey: GenAI Public Cloud Spend To Grow 4X '25-'28, TD Cowen
