Viaduct

High-Level Overview

Viaduct is a Menlo Park-based AI software company that builds a cloud platform analyzing connected asset data to reduce warranty costs, downtime, and improve customer experiences for manufacturers in automotive, trucking, construction, agriculture, and industrial sectors.[1][2][3][4] Its core product uses patented unsupervised AI to detect anomalies in complex time-series data, enabling predictive maintenance, quality issue identification, and operational optimization—such as cutting warranty expenses by 12% on average and unplanned downtime by 23%.[2][4] Founded in 2018, Viaduct has raised $21M in funding (latest $10M Series B round about a year ago) and serves enterprise clients by compressing and enhancing vehicle data for cloud insights, powering use cases like driver personalization and safety scores.[2][3] As of August 2025, Sumitomo Rubber Industries acquired Viaduct to bolster its AI capabilities.[5]

Origin Story

Viaduct was founded in 2018 by a team of experts in engineering, AI, product management, and business development, led by CEO David Hallac.[1][2] Key leaders include Ken Anderson (Head of Engineering), Brian Belloli (Chief Commercial Officer), Jim Brady (VP of Business Development), Matej Drev (COO), and Matthew Howard (Chief Product Officer), whose combined backgrounds shaped its focus on enterprise-grade AI for data-heavy industries.[1] The idea emerged from recognizing untapped potential in connected vehicle and asset data, evolving from general AI software to a specialized platform for predictive analytics amid rising IoT adoption in manufacturing.[2][3][4] Early traction built through deployments in fleet maintenance and automotive, culminating in Series B funding and the 2025 acquisition by Sumitomo Rubber, marking a pivotal shift to scaled global impact.[2][5]

Core Differentiators

• Patented Unsupervised AI for Time-Series Data: Unlike traditional methods, Viaduct's tech automatically discovers hidden patterns and anomalies in massive connected asset datasets weeks or months earlier, isolating issues by factors like build line or environment without predefined models.[1][3][4]
• Edge-to-Cloud Data Optimization: In-vehicle machine learning compresses and enriches data for efficient transmission, enabling real-time insights while minimizing bandwidth—key for scalable applications like predictive maintenance and driver safety.[2][3]
• Unified Platform with Fast Query Engine: Provides a single view of structured asset data for technical and business users, slashing investigation time by 90% and delivering actionable predictions for warranty reduction and downtime prevention.[4]
• Industry-Specific Focus: Purpose-built for connected assets in automotive and beyond, outperforming general analytics in fleet maintenance (named a Challenger by CB Insights alongside Samsara).[2]

Role in the Broader Tech Landscape

Viaduct rides the explosion of connected assets and IoT, where manufacturers generate petabytes of time-series data but struggle with actionable insights amid rising warranty pressures and downtime costs.[2][4] Timing aligns with automotive electrification, autonomous trends, and predictive maintenance demands—exacerbated by supply chain disruptions—positioning its AI as essential for incumbents like automakers optimizing fleets.[2] Market forces like AI democratization and edge computing favor Viaduct, enabling revenue from premium services (e.g., personalization) while cutting costs; its acquisition by Sumitomo Rubber amplifies influence in tire and mobility ecosystems, accelerating AI adoption across heavy industry.[5] In the fleet maintenance landscape, it challenges leaders by transforming raw telematics into business outcomes, influencing how operators shift from reactive to proactive strategies.[2]

Quick Take & Future Outlook

Post-acquisition by Sumitomo Rubber in August 2025, Viaduct will likely integrate deeply into mobility giants' operations, expanding its AI from analytics to end-to-end manufacturing optimization and premium IoT services.[5] Trends like generative AI for simulations, stricter emissions regs, and autonomous fleets will propel demand, potentially evolving its platform toward multi-asset orchestration (e.g., vehicles + tires + infrastructure). Expect scaled deployments reducing industry-wide downtime further, cementing Viaduct's role in data-driven manufacturing—transforming raw signals into sustained competitive edges, much like its origin in unlocking hidden patterns from connected chaos.[4]