October 2025 Summaries

Chalk has been recognized by Fast Company as one of the "Next Big Things in Tech 2025," highlighting its innovative approach to AI infrastructure. The annual list identifies technologies with the potential to transform industries within five years, and Chalk's solution addresses the traditional tradeoff in machine learning between speed and data freshness. By computing features at inference time, Chalk enables real-time data access without the need for ETL jobs, offering improved performance and ease of use. This approach is particularly beneficial for critical machine learning applications like identity verification and fraud detection, with companies such as Whatnot, MoneyLion, and Socure utilizing Chalk to scale their ML operations. The company's recent $50 million Series A funding, led by Felicis, underscores its commitment to advancing real-time ML infrastructure, positioning it alongside other innovative companies like Runway, GitHub, and Roblox.

Chalk is designed to address the needs of enterprises building systems such as fraud detection, recommendation engines, or search functionalities by running entirely within a customer's cloud infrastructure, specifically inside their VPC and on Kubernetes, to ensure data security, predictable latency, and internal control. By deploying in the customer's account, Chalk offers a strategic advantage by maintaining data in-account, thus enabling compliance, auditing, and governance to be managed internally. The platform is built to deliver low latency for real-time decisions by co-locating feature computation with applications, which reduces cross-cloud data transfer and ensures consistency across online and offline pipelines. Additionally, Chalk's integration with Kubernetes allows for uniform deployments and controlled upgrades, providing scalability and operational efficiency without egress overheads. Andrew Moreland, co-founder of Chalk, emphasizes these strategic differentiators on the Adventures in DevOps podcast, highlighting how Chalk ensures fast and accurate machine learning feature computation by leveraging temporal data awareness and seamless management of IAM.