Kaskada

Kaskada, based in Seattle, Washington, provides a machine learning platform for feature engineering using event-based data, enabling data scientists and engineers to collaborate on creating and sharing features for ML models. The company secured $9.8 million in total funding, including a $1.8 million seed round in 2018 and an $8 million Series A in 2020, and reported 10-20 employees at the time of its Series A. Its platform processes massive event streams or databases to generate time-based features for real-time predictions, unifying feature creation and serving to streamline data workflows for AI applications. The company's core technology was later acquired by DataStax, which subsequently planned to open-source it. Key investors included Voyager Capital, NextGen Venture Partners, and Founders’ Co-op. Kaskada was founded in 2018 by Davor Bonaci and Ben Chambers.

High-Level Overview

Kaskada is a Seattle-based machine learning startup founded by former Google Cloud engineers, focused on building a software platform that enables seamless collaboration between data scientists and data engineers.[4][5] It serves enterprises handling big data, solving the problem of siloed machine learning workflows by providing tools for real-time feature engineering and model development without traditional infrastructure hurdles.[4][5] The company raised $8 million in Series A funding in 2020, bringing total funding to $9.8 million, signaling early growth momentum in the competitive ML operations space.[5]

Origin Story

Kaskada emerged in 2020 from the expertise of ex-Google Cloud engineers who identified pain points in machine learning pipelines during their time at Google.[4] The founders recognized that data scientists and engineers often struggled with collaboration due to fragmented tools and data pipelines, leading to the idea for a unified platform.[5] Early traction came swiftly with the $8 million Series A round announced shortly after launch, backed by investors betting on its potential to streamline ML feature stores and real-time processing.[4][5]

Core Differentiators

• Unified ML Platform: Unlike traditional systems requiring separate data engineering and science tools, Kaskada's software allows collaborative building of machine learning models directly on raw data sources, emphasizing real-time feature computation.[4][5]
• Developer Experience: Simplifies workflows for big data environments, reducing setup time and enabling engineers and scientists to work together without custom infrastructure.[4]
• Scalability and Speed: Handles large-scale data processing efficiently, positioning it as a feature store alternative focused on collaboration over standalone storage.[5]

Role in the Broader Tech Landscape

Kaskada rides the wave of the MLOps (Machine Learning Operations) trend, where enterprises demand faster, more reliable AI deployment amid exploding data volumes.[4] Its timing aligns with the post-2020 surge in real-time ML needs driven by cloud-native apps and edge computing, amplified by market forces like talent shortages in data teams and the shift from batch to streaming analytics.[5] By fostering cross-team collaboration, Kaskada influences the ecosystem toward democratized ML, potentially accelerating adoption in sectors like finance, e-commerce, and healthcare where predictive models drive competitive edges.

Quick Take & Future Outlook

Kaskada is poised for expansion in the maturing MLOps market, likely pursuing Series B funding and enterprise partnerships to scale its platform amid AI infrastructure consolidation.[4][5] Trends like generative AI and agentic systems will amplify demand for its real-time collaboration tools, evolving its role from niche startup to key player in unified data platforms. As ML becomes table stakes for tech companies, Kaskada's focus on humanizing complex workflows could solidify its influence, transforming how teams build production-grade models.