Participated · Seed · Aug 2026

Riven Systems was founded to rethink industrial chemistry design using density functional theory (DFT), AI, and autonomous experimentation to design highly selective ligands for separations in critical minerals processing. The team includes co-founders Orion Archer Cohen and Kurt (Chip) Breitenkamp and has built an autonomous lab in New York City. Riven says it is hiring across computational, experimental, and engineering sciences to grow its founding team. Financially, the company announced a $7.9 million financing led by Cerberus Ventures, participation from several VC firms, and a DOE Genesis Mission grant. New System Ventures incubated the company and acted as an early backer through its partners. Riven positions itself to address bottlenecks in critical minerals supply chains by accelerating development and deployment of chemistry solutions.

Participated · Seed · Jul 2026

Piq Energy builds an agentic grid planning platform that sits on top of industry engineering tools to connect models, data and workflows and automate complex engineering studies. The company aims to streamline interconnection processes by enabling agents to run studies simultaneously, unlock existing grid capacity, and accelerate adoption of flexible transmission solutions. Piq launched in 2025 after stealth development and reports managing more than 1,000 transmission grid models and completing over 10,000 engineering workflows while supporting customers running dozens of analyses daily. The company was founded in 2023 and is headquartered in San Francisco. Its founders, Tom Nudell, PhD (CEO), and Dionysios Stamatiadis, MSc (Chief Engineer), bring experience from utilities and technology firms including Smart Wires, UK Power Networks, EirGrid, APS, Oncor, Palantir and C3 AI. The recent $5 million seed raise led by Active Impact Investments will fund the next phase of product rollout to utilities and developers and further development of Piq’s grid intelligence layer.

Participated · Seed · May 2025

Rhizome provides a cloud‑based climate resilience planning platform for utilities, built on its scalable Aspen platform. Its flagship product, gridADAPT, supports long‑term infrastructure planning and prioritizes resilience investments; it has since launched gridFIRM for wildfire risk mitigation and gridCAVA for climate vulnerability assessments for municipal and cooperative utilities. The company leverages AI applied to climate risk data and digital representations of physical grid assets to identify vulnerabilities and prioritize upgrades. Current customers include Avangrid, Seattle City Light, Vermont Electric Power Company (VELCO) and Vermont Electric Cooperative (VEC). Rhizome also has strategic collaborations with EPRI’s Climate READi, KPMG, Black & Veatch, and the University of Connecticut. The company intends to use the funds to scale its AI platform and expand its team to help utilities protect their grids and customers from extreme weather impacts. Rhizome provides an AI-powered climate resilience planning platform called gridFIRM that helps utilities identify vulnerabilities from climate threats and quantify risk at high resolution. The platform applies machine learning to asset, geographic, and historical datasets to deliver granular, asset-level insight, including the likelihood an asset could ignite a wildfire. It also quantifies the economic and social benefits of grid-enhancing investments to support utility planning and wildfire mitigation. gridFIRM is available to electric utilities in the U.S., Canada, and Australia and was developed using feedback from customers in the Northeast, Pacific Northwest, and Texas. Led by CEO Mishal Thadani and CTO Rahul Dubey, Rhizome plans to use the $1M investment from Convective Capital to expand operations and advance product development. The company is headquartered in Washington, D.C.

Participated · Series A · Sep 2024

Brightband is developing machine-learning models trained on years of observational weather data to deliver faster, cheaper weather and climate forecasts than traditional physics-based systems. The company plans to ingest and release large, underused datasets (satellite and weather‑balloon observations) to improve model performance. Brightband aims to open source its core forecasting models, the training data, and evaluation metrics while offering paid, specialized services on top. The team is building on existing physics-based models and research, targeting state-of-the-art accuracy and much lower compute and latency. Use cases highlighted include renewables forecasting for energy, transportation risk avoidance, and agricultural planning. The product is early — nothing is live today — and Brightband expects a model that ingests observations and produces forecasts by the end of 2024. The company is organized as a public benefit corporation and has raised $10 million in funding.