Accelerating Battery Innovation: How Holyvolt Uses AI to Challenge China’s Supply Chain Dominance

• Holyvolt secured $45 million in funding, valuing the company at $245 million, to accelerate the discovery of Western battery materials using AI.
• The company’s Discovery Engine integrates proprietary AI models with automated labs built on Wildcat Discovery Technologies’ high-throughput platform, following a 2026 merger.
• Holyvolt possesses an 18-year proprietary dataset comprising over 500,000 physical battery cells, including failed experiments that are typically excluded from standard AI training.
• The firm’s first commercial product, DRX, is a nickel- and cobalt-free cathode designed to use raw materials sourced outside China, with commercialization planned for 2027.
• New US Pentagon regulations taking effect on October 1, 2027, will prohibit the procurement of batteries from CATL, BYD, and four other Chinese manufacturers, creating an immediate defense sector opportunity.
• Holyvolt’s high-throughput combinatorial chemistry lab tests thousands of combinations in parallel, delivering results ten times faster than conventional R&D and potentially accelerating development by 100-fold when combined with AI.

Holyvolt’s approach to solving the battery bottleneck relies on a closed-loop system known as the Discovery Engine. This platform connects proprietary AI models to automated laboratories that physically build and test real battery cells. As Steven Prast, Chief Operating Officer (COO), explains, each round of physical results improves the next AI prediction. The AI identifies how variations in materials and production conditions affect performance, allowing the firm to optimize these variables before committing to expensive manufacturing trials.

The infrastructure for this engine is built on a high-throughput platform developed by Wildcat Discovery Technologies, which Holyvolt merged with in 2026. This merger provides the physical testing capabilities necessary to validate AI predictions. Mathias Ingvarsson, Founder and CEO of Holyvolt, emphasizes that traditional R&D in battery chemistry is inefficient because it involves millions of possible material combinations. While traditional methods take years to identify viable options, Holyvolt’s AI analyzes enormous amounts of experimental data to predict promising combinations and determine which experiments to run next.

Crucially, the system relies on reinforcement learning. After the AI predicts a combination, physical tests are conducted in the lab. The real-world data from these tests is fed back into the AI model, making the next cycle more accurate and efficient. Ingvarsson notes that Holyvolt has a significant competitive advantage due to its history: "We've spent 18 years building a proprietary dataset covering more than 500,000 physical battery cells, including failed experiments that most AI models never see." This comprehensive data allows the platform to deliver results ten times faster than conventional R&D.

The speed of discovery has profound implications for commercialization. Ingvarsson states that the true value of AI in this sector will be measured by how quickly better batteries can be integrated into real products. One of the primary challenges in battery innovation is the time required to transition a promising material from the lab to a commercially viable product. A chemistry may perform well in controlled environments, but scaling it requires extensive testing for performance consistency, cost, and production requirements.

Holyvolt applies the same high-throughput experimentation approach that transformed drug discovery to battery development. By building and testing thousands of material combinations in parallel rather than sequentially, the firm can evaluate more materials and identify manufacturing challenges earlier. This allows Holyvolt to focus resources on chemistries with the greatest potential to scale, addressing costly limitations years before they would typically be discovered.

The first material from Holyvolt moving toward market is DRX, a nickel- and cobalt-free cathode designed to deliver high energy density using raw materials sourced entirely outside China. The company plans to commercialize DRX in 2027. This timeline aligns with new US defense regulations; starting October 1, 2027, the Pentagon will no longer be able to procure batteries from CATL, BYD, and four other Chinese manufacturers. As the US military expands its use of drones, Holyvolt intends to target this sector first as these restrictions take effect.

Steven Prast, formerly a Senior Engineering Manager at Tesla, brings extensive experience from the Elon Musk-founded EV company to Holyvolt’s operations. Prast highlights that developing breakthrough technology is only half the battle; the real challenge lies in manufacturing it consistently, cost-effectively, and at the scale required for meaningful impact. His insights inform Holyvolt’s strategy to streamline the transition from lab discovery to factory manufacturing.

The combination of Holyvolt’s proprietary data, high-throughput physical testing, and AI-driven prediction creates a reinforcement learning cycle that grows more powerful with every experiment. Ingvarsson asserts that this infrastructure positions the US and Europe to leapfrog China’s decades-long consolidation of the battery supply chain. By discovering new chemistries and developing efficient manufacturing methods, Holyvolt aims to reduce dependence on Chinese supplies while accelerating the next generation of battery innovation. The company’s investors in this latest round include Fifth Wall, Zone II, and CC.VC, signaling strong confidence in its ability to alter the commercial timeline for battery chemistry.