Startups & Funding
Sift raises $42M to scale manufacturing data infrastructure
Sift, a data infrastructure startup for complex manufacturing, raised a $42 million Series B at a $274 million post-money valuation to support AI-driven factory automation.
Sift, a California-based data infrastructure company founded by former SpaceX engineers, has raised a $42 million Series B funding round. The investment, which closed in 2025, values the company at a $274 million post-money valuation. Founded in 2022 by CEO Karthik Gollapudi and CTO Austin Spiegel, the El Segundo-based startup plans to use the capital to expand its platform, which helps manufacturers manage telemetry data. Telemetry data refers to real-time performance information streamed from sensors on physical components during testing, manufacturing, and launch.
The funding round was led by StepStone, with participation from:
- GV (Google’s venture capital arm)
- Riot Ventures
- Fika Ventures
- CIV
Sift provides software tools to organize and store telemetry data for companies building complex machines, such as spacecraft and cars. Its customer base includes U.S. rocket builder United Launch Alliance and satellite manufacturer Astranis. Managing these data streams is highly resource-intensive; according to Sift, some vehicles it works with have more than 1.5 million sensors streaming data concurrently across multiple formats and time scales. For companies like Astranis, which might perform 10 million automated software tests in a day, the volume of data can quickly become a financial burden. Jeff Dexter, the VP of software at Astranis, noted that without proper infrastructure, “Inevitably, it gets to a point where it’s costing us millions of dollars per month just to store data.” Dexter added that Sift’s technology helps alleviate concerns over data volume and storage costs, allowing the company to focus on utilizing its data effectively.
The startup is shifting its focus to ensure this machine data is structured for artificial intelligence and deep learning models. According to Gollapudi, the company’s goal is to make complex telemetry data machine-readable so AI agents can analyze test data and make manufacturing decisions. Gollapudi noted that the company’s long-term vision for how this transition would play out over five years is instead occurring this year. This pivot comes amid a broader industry push toward physical manufacturing—a concept often referred to as “atoms, not bits” to contrast it with digital products—and factory automation. This trend was highlighted last week by reports that Jeff Bezos is creating a $100 billion fund to automate factories.
Why it matters
Sift is pivoting its data infrastructure focus to support AI and deep learning models, aiming to make complex machine data machine-readable for automated manufacturing. As physical industries increasingly rely on software-intensive operations, structuring massive sensor streams becomes essential for deploying AI decision-making tools on the factory floor.