Participated · Series B · Jul 2026

CuspAI builds AI systems that search for and design new materials tailored to specified properties, serving customers including ASML and Meta. The company counts AI researchers Yann LeCun and Geoffrey Hinton as advisers. It plans to expand geographically across the US, APAC and Europe and to deploy an “AI Materials Foundry,” a global network of data, labs, compute and scientific expertise with founding partners such as Nvidia, Meta, Samsung and Hyundai and more than 40 other companies. Financially, CuspAI is a two-year-old startup that has raised over $670M to date, including a $450M Series B that values it at $2.6B. The firm says its platform addresses materials discovery bottlenecks across industries including energy, semiconductors, manufacturing, water treatment and carbon capture.

Led · Series B · Mar 2026

Sift provides a unified data platform that ingests audio, video, logs, and high-frequency telemetry and automatically transforms raw sensor data into structured, queryable formats for engineers and AI systems. The platform is used as a system of record by customers including ULA, Astranis, K2 Space, Parallel Systems, and undisclosed enterprise defense programs to diagnose anomalies, validate designs, and maintain fleet-wide visibility. Founded by SpaceX alumni Karthik Gollapudi and Austin Spiegel, Sift aims to enable fleet-scale operations by automating anomaly detection and providing unified observability across thousands of machines. The company has about 70 employees and plans to nearly double its team, relocate to a larger Marina Del Rey headquarters, and expand platform capabilities. Financially, Sift has raised a total of $67 million following its $42 million Series B.

Led · Series B · Mar 2026

Founded in 2022 and based in El Segundo, California, Sift Stack develops data infrastructure and software to manage massive telemetry streams from complex machines such as rockets, satellites, and vehicles. Its tools organize and store sensor data—some vehicles stream over 1.5 million concurrent sensors across multiple formats and timescales—to make that information machine-readable for AI and automation workflows. Customers include United Launch Alliance, Astranis, defense contractors, robotics companies, and power-grid management startups. The founders, CEO Karthik Gollapudi and CTO Austin Spiegel, previously worked on telemetry tooling at SpaceX and built Sift to offer a best-in-class alternative to ad hoc database and script-based approaches. The company says the arrival of AI tools has accelerated demand for robust data infrastructure, prompting a shift in its focus over the past year. Financially, Sift Stack raised a $42 million Series B in 2025 at a $274 million post-money valuation led by StepStone with participation from GV, Riot Ventures, Fika Ventures, and CIV.

Participated · Series A · Nov 2025

Exowatt’s core product is the P3 unit, a shipping-container-sized metal box topped with lenses that concentrate sunlight onto heat-retaining bricks. A fan carries the heat to a Stirling engine and generator, enabling each thermal battery to store energy for up to five days and deliver round-the-clock electricity. The system’s efficiency rivals photovoltaic solar plus lithium-ion storage, while its compact, modular design is intended to unlock mass manufacturing learning curves. Exowatt is targeting production of one million units per year, a scale that management says will let it hit its one-cent-per-kilowatt-hour cost goal. Demand appears strong: the company reports a backlog of roughly 10 million P3 units, equivalent to 90 GWh of capacity. Longer term, Exowatt intends to deploy millions—and eventually billions—of units to power the rapidly expanding fleet of AI data centers in sun-rich regions. The approach repurposes decades-old concentrated solar technology but packages it for low-cost, large-scale manufacturing.

Led · Series C · Nov 2025

Founded in 2019, Scribe makes two core products: Scribe Capture, which records user actions to generate step-by-step guides with text and screenshots, and the newly announced Scribe Optimize, which mines activity data across an enterprise to surface the workflows where automation or AI will deliver the greatest return. The platform has documented more than 10 million workflows spanning 40,000 software applications and now serves over 5 million users. Adoption is broad—teams inside 94 % of the Fortune 500 use Scribe, and 78,000 organizations are paying customers, including New York Life, T-Mobile, LinkedIn, HubSpot, and Northern Trust. Users report saving 35–42 hours per person each month and onboarding new hires 40 % faster. Revenue more than doubled in the past year, although specific figures were not disclosed, and the company’s valuation has grown five-fold since its prior round to reach $1.3 billion. Headcount stands at 120 employees, with plans to double over the next 12 months. The new capital will accelerate the rollout of Scribe Optimize and related products as enterprises scramble to decide where to deploy automation and generative AI.