Feldera Raises $21.5M to Reduce Database Compute Costs by 95% with a Mathematical Breakthrough
Feldera, the incremental view maintenance engine for enterprises, led by Lalith Suresh, Leonid Ryzhyk, Mihai Budiu, Ben Pfaff, and Gerd Zellweger, has raised $21.5 million across a $15.4M Series A round led by Inovia Capital, with participation from Costanoa Ventures and Battery Ventures, and a $6.1M Seed round led by Costanoa Ventures, with participation from Ion Stoica (co-founder of Databricks and Anyscale).
Founded by award-winning computer scientists, Feldera has built an engine that makes it dramatically faster and more cost-efficient to keep answers fresh across massive datasets, for both AI agents and the people relying on them.
Enterprises’ Problem: AI Agents Need Fresh Data to Deliver AI-Ready Products
Companies today hold more data than ever, yet turning that data into a reliable foundation for AI agents remains expensive and slow. Legacy analytics platforms often rely on brute-force execution, repeatedly scanning entire databases whenever a query runs, even when only a small fraction of the underlying data has changed.
This computational redundancy creates a difficult trade-off for AI-ready products: absorb rapidly rising compute costs or settle for stale, shallow data that limits AI performance and slows operational decision-making.
Feldera’s Solution: Incremental View Maintenance to Reduce Compute Time & Costs.
Feldera addresses this bottleneck through incremental view maintenance, a technology that keeps query results continuously up to date across massive enterprise databases. Built on DBSP (Database Stream Processing), the underlying theory automatically converts complex SQL queries into incremental programs that process only newly changed data, eliminating redundant full-database scans.
The platform integrates directly with existing data lakes without requiring major architectural changes. According to Feldera, this can reduce compute costs by 95% or more while transforming hours-long analytical pipelines into sub-second updates.
The result is continuous, low-latency access to accurate enterprise context for AI agents, without repeatedly consuming cloud compute on unchanged data. Feldera says customers are already running the technology in production for use cases including real-time fraud detection, large-scale logistics operations, and fine-grained authorization for AI agents.
Key features include:
Universal SQL Incrementalization: Converts complex SQL programs into incremental programs without falling back to full recomputation.
Seamless Integration: Connects directly with existing enterprise data warehouses, data lakes, and pipelines without requiring custom engineering or brittle streaming code.
No Rewrites Required: Enables enterprises to migrate existing SQL workloads to Feldera without code changes, potentially replacing months of migration work.
A Fraction of the Hardware: Feldera reports that customers have reduced infrastructure costs by an average of 10×, with some workloads achieving savings of more than 100×.
Enterprise Scale, Laptop Simplicity: Processes millions of records per second, enabling workloads that once required a cluster to run on a single laptop.
Mathematically Guaranteed Correctness: Backed by DBSP, the underlying theory provides formal guarantees that Feldera’s results match those produced by the corresponding warehouse, regardless of SQL complexity.
"Making timely decisions against massive amounts of fast-changing data using traditional methods requires an immense amount of compute," said Lalith Suresh, CEO and co-founder of Feldera. "AI agents are only as good as the data they can access. So if that data is too stale because it is too expensive to compute, companies will burn tokens on agents that are just going to be wrong all the time."
World-Class Research Team Behind 200+ Papers.
Feldera was founded in 2023 by five former VMware researchers, Lalith Suresh, Leonid Ryzhyk, Mihai Budiu, Ben Pfaff, and Gerd Zellweger. Collectively, the team has published more than 200 research papers spanning database systems, distributed computing, and operating systems.
The team’s work grew out of a recurring problem: engineers were relying on fragile, ad-hoc solutions to handle incremental data processing. At VMware Research, the researchers tackled the problem at its mathematical foundation, leading to DBSP, an award-winning framework for incremental SQL computation.
Feldera was created to bring this research into production, making advanced incremental computation accessible to enterprise engineering teams. The founding team combines academic and technical backgrounds from institutions including Stanford, Carnegie Mellon, ETH Zurich, UNSW, and TU Berlin, alongside decades of experience designing large-scale systems.
The result is an enterprise-ready engine built on decades of database research, bringing previously difficult incremental computation capabilities into modern data infrastructure.
"The team at Feldera solved a 50-year-old fundamental database problem with rigorous mathematical proof," said Taha Mubashir, Partner at Inovia Capital. "As enterprises scale their real-time AI initiatives, the cost and latency of traditional data compute become unsustainable. Feldera provides the foundational compute layer that enables companies to run continuous, complex analytics at a fraction of the cost."
𝐑𝐞𝐦𝐨𝐭𝐞 𝐇𝐢𝐫𝐞 𝐨𝐫 𝐁𝐮𝐢𝐥𝐝 𝐖𝐢𝐭𝐡 𝐔𝐬: 𝐁𝐮𝐢𝐥𝐝𝐰𝐞𝐫𝐤𝐬 helps technology companies build in India: two ways. Hire vetted remote engineers, product leaders, designers, and GTM professionals directly onto your team. Or partner with a proven development studio to design, build, and ship your product end-to-end. Learn More At: menlotimes.com/buildwerks
