Participated · Seed · Jun 2026

Atana Elements uses proprietary AI and machine-learning models combined with deep subsurface expertise to identify and de-risk overlooked mineral deposits. The company claims it can accelerate discovery timelines to roughly 22 months, compared with the typical five-to-ten year cadence. Atana is currently focused on lithium exploration, with plans to expand into helium, hydrogen, and copper. It already has projects underway across the EU and the Americas. The company is led by CEO Thomas Wilson. Financially, Atana recently closed a $27.5 million seed round from Overture Ventures to support its ongoing exploration and growth activities.

Participated · Equity · Mar 2026

Founded in 2023, Archimetis creates an artificial-intelligence decision-making system designed for refineries and other heavy-industry plants. The software aggregates scattered operational data and delivers real-time recommendations, helping plant teams overcome complex production decisions. By targeting the energy and chemical sectors, the company aims to unlock significant efficiency gains and cost savings. Archimetis differentiates itself through the Google-honed AI expertise of founder Paul Manwell. The fresh capital will be used to enhance the product and expand deployments across additional industrial facilities. No revenue, customer count, or other operating metrics were disclosed in the article.

Participated · Equity · Nov 2025

Everstar is a U.S.–based startup behind Gordian, an artificial-intelligence platform that automates regulatory compliance and licensing processes for nuclear plants and other critical industrial operations. Gordian combines AI with deep sector expertise to prepare technical documentation, streamline licensing workflows, and reduce compliance costs while strengthening collaboration between companies and regulators. Its capabilities cover both existing plants and new reactor projects, including small modular reactors (SMRs). The platform has already cut Last Energy’s expected eight-week environmental assessment to just one week, demonstrating meaningful time savings. Everstar sees growing demand for its solution as AI proliferation and data-center expansion drive global energy needs and heighten interest in low-carbon, always-on nuclear power. The company recently secured $5.5 million in new funding, adding to a pre-seed round completed in early 2025.

Participated · Series A · Apr 2025

Fourier designs hydrogen electrolyzer modules no larger than two standard server racks, using arrays of small "blade" electrolyzers managed by software. Each module contains about 20 blades fed by a shared pump and powered by lightly modified, commodity data‑center power supplies. The company’s control software monitors individual blades similar to a battery management system, optimizing output and detecting degradation. Fourier has run two lab‑scale pilots that produce about one kilogram of hydrogen per hour with a pharmaceutical manufacturer and a solar energy company. It is deploying two commercial‑scale pilot plants—one at a petrochemical plant in Ohio and another at a Fremont, California aerospace parts company—expected to be operating by June. Fourier targets customers that need 6–20 kilograms per hour (roughly 300 kW to 1 MW of capacity). The company says it can deliver hydrogen for $6–$7 per kilogram (ex‑incentives) versus customer current prices of about $13–$14 per kilogram.

Participated · Series A · Feb 2025

Tyba provides a unified simulation and operations platform for energy storage that combines AI-enabled forecasting, automated dispatch and bidding strategies with project simulation tools. Its Asset Operations product automates execution while letting operators make quick strategic adjustments, and its Project Simulation product has been used to model over 100GW of projects under development. Tyba supports operations of more than 1 GWh of storage assets in Texas and California, has tripled its customer base in the past year, and delivers revenue outcomes in the top 5% of assets. Since founding, Tyba’s simulation work has helped secure over $1 billion in project funding. The company is hiring across engineering, modeling, and commercial functions and plans to expand its forecasting and optimization capabilities into new markets and across asset classes. Tyba is headquartered in San Francisco. Tyba provides an accessible modeling platform that simplifies and unifies project modeling applications for solar and storage development, financing, and operations. The platform is used by leading community and utility-scale solar and storage firms to determine solar production, inform energy storage dispatch, leverage historical and forecasted wholesale price data, and generate risk-adjusted bids for physical power delivery. Tyba combines AI/ML for price forecasting, cutting-edge physical models for energy production, and linear optimizers to drive valuation and dispatch decisions. The company is led by Michael Baker (CEO), Tyler Nisonoff (CTO), and Tom Thunell (COO). Tyba raised $2.25M in seed funding and intends to use the proceeds to expand operations and accelerate development efforts. The platform’s focus is on improving modeling accuracy and usability across development and operations workflows.