Eventual: Revolutionizing Multimodal Data Infrastructure for AI Applications
Founders Sammy Sidhu and Jay Chia, former software engineers at Lyft’s autonomous vehicle program, identified a critical challenge early on: managing the massive influx of unstructured data generated by self-driving cars. From 3D scans and photos to audio and text, the data was overwhelming, and existing tools couldn’t efficiently process or analyze all these diverse data types in one unified platform. Engineers faced a patchwork of open-source solutions that were unreliable and time-consuming, hampering innovation.
Sidhu recalls, “We had talented PhDs and experts working on autonomous vehicles, but they were spending up to 80% of their time on infrastructure rather than on developing the core technology.” This realization sparked the creation of an internal multimodal data processing tool at Lyft. As Sidhu explored new career opportunities, he kept hearing from interviewers about the need for similar data solutions, inspiring him to launch Eventual.
Eventual has developed Daft, an open-source, Python-native data processing engine designed to handle diverse data modalities—from text and audio to video—quickly and efficiently. Sidhu envisions Daft becoming as transformative for unstructured data as SQL was for structured, tabular datasets.
Founded in early 2022—before the release of ChatGPT and widespread awareness of the data infrastructure gap—Eventual launched the first version of Daft later that year. The company is now preparing to introduce an enterprise version in the third quarter, aiming to serve larger organizations with robust, scalable solutions.
The surge in AI applications, especially following the popularity of ChatGPT, has accelerated demand for multimodal data processing. Companies are increasingly integrating images, videos, and documents into their AI models, expanding the need for flexible, high-performance data infrastructure. As Sidhu notes, “Usage of multimodal data has skyrocketed, and we’re seeing a huge uptick in how organizations handle complex, unstructured information.”
While Eventual’s roots are in autonomous vehicles, its technology has broad applications across industries like robotics, retail tech, and healthcare. Major clients include Amazon, CloudKitchens, and Together AI, among others.
The company’s recent funding rounds—$7.5 million in seed funding led by CRV and a subsequent $20 million Series A led by Felicis with participation from Microsoft’s M12 and Citi—are fueling its growth. Funds will be used to expand Daft’s open-source capabilities and develop a commercial product that enables clients to build sophisticated AI applications using processed multimodal data.
Felicis partner Astasia Myers highlighted Eventual’s unique position, noting that it emerged as a first mover in a rapidly evolving space. With the multimodal AI industry projected to grow at a 35% compound annual rate through 2028, the importance of efficient, native data processing engines like Daft will only increase. Myers emphasizes that with 90% of data being unstructured—and global data generation increasing exponentially—Eventual’s solutions are poised to meet a critical, growing need in the AI ecosystem.
