How was this content?
Announcement: the Compute for Climate Fellowship’s 2025 cohort is here
Meet the 23 changemakers shaping a more sustainable future on land, at sea, and in the air
Global climate action demands bold ideas and innovative technology. It also demands collaboration across both public and private sectors, industries, governments, and research leaders. Appetite for change is widespread, and as technologies like generative AI are opening new possibilities for breakthrough progress, now is the time to invest in novel, scalable, and sustainable climate solutions. That’s where the Compute for Climate Fellowship can help.
The Compute for Climate Fellowship is a global R&D funding program founded in 2023 by Amazon Web Services (AWS) and the International Research Centre on Artificial Intelligence (IRCAI). Every year, a select group of companies are accepted into the program and empowered to turn bold ideas into climate solutions that deliver real results.
A fellowship built on innovation and a shared responsibility to power change
Members of the fellowship develop groundbreaking proof-of-concept (POC) climate solutions. These solutions must deliver value across at least one of these nine key areas: clean energy, low-carbon transportation, sustainable agriculture and food, circular economy and industry, sustainable buildings, greenhouse gas (GHG) management, carbon removal, environmental and climate risk, and indigenous solutions to the climate crisis.
Once accepted, members receive grants in the form of AWS Credits as well as access to leading scientific and technical support from AWS and IRCAI. The build process runs for three months and participants get personalized technical mentorship from AWS and IRCAI experts every step of the way. They can also access advanced computing services from AWS, including high-performance compute (HPC), quantum computing, and generative AI. For selected participants, the Fellowship also provides media and promotional opportunities, including showcasing POC results across United Nations and other international bodies.
Ramping up investment and accelerating progress
Acute climate challenges are being felt across the globe and AWS and IRCAI are doubling down our efforts in response. This year, we have expanded the program’s size to offer up to $4 million USD in total funding available for 23 companies. That’s up from $1.5 million USD for 8 companies in 2024.
This year, the Fellowship is also pursuing a new focus on solutions that are built by, with, and for indigenous communities. These communities are often directly impacted by climate change and are important leaders in addressing the challenges we all face. They contribute invaluable knowledge, innovations, and practices that will help us build powerful solutions to the climate crisis and biodiversity loss.
Now, without further ado, let’s say hello to this year’s cohort.
80 Acres Farms is a leader in vertical farming, increasing sustainable food production in areas where drought is common. The company operates indoor farms with 100 percent renewable electricity and 95 percent less water usage by combining class-leading technology and deep analytics. As part of its work with the Compute for Climate Fellowship, 80 Acres Farms is building a machine learning-powered genetic computing platform capable of accelerating precision breeding and improving genetic crop optimization.
Aionics is accelerating the discovery of next-generation battery materials using AI. Its platform, built by experts from Stanford and the University of Michigan, enables rapid simulation and screening of novel electrolytes and components. As part of the Fellowship, Aionics is building computational models for non-flammable lithium-ion battery electrolytes using machine learning (ML)-powered interatomic potentials and quantum simulations on AWS.
Amini is a deep tech startup focused on transforming climate resilience in the Global South through advanced satellite and ground data integration. With a mission to close data and compute gaps, Amini powers insights for over 300,000 smallholder farmers. In the Fellowship, it is developing a generative AI foundation model for land-based climate intelligence in Barbados, leveraging multi-temporal, multispectral satellite data.
Anthrogen engineers novel enzymes for carbon capture and conversion, using cutting-edge Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and synthetic biology. Their technology enables CO₂ transformation into valuable polymers and fuels. As part of the Fellowship, Anthrogen is developing a generative AI foundation model as part of a discrete diffusion framework capable of simulating billions of years of protein evolution. These simulations will then be used to drive the development of new-to-nature enzymes for clean manufacturing.
Aura Aero is a French aerospace manufacturer developing electric and hybrid aircraft, including a 19-seat electrical regional aircraft (ERA) commuter plane. The company is committed to helping the aviation industry reduce emissions by 55 percent by 2030 and reach carbon neutrality by 2050. As part of that effort and their participation in the Fellowship, the Aura Aero team is building a high-performance computing farm to conduct fully digital simulations of aircraft aerodynamics, structure, and energy systems in the cloud.
Avalo uses machine learning and evolutionary biology to accelerate crop development. The company’s rapid evolution platform accelerates genetic recombination to create resilient, high-performing plants. Avalo is now building an AI-driven genomic selection system using simulations to optimize and improve crop yields.
Barnwell Bio is improving animal health through biosurveillance and metagenomic sequencing, with a focus on commercial farming. As part of the Fellowship, the company is developing MetaVet, an AI chatbot that combines large language models (LLMs) with predictive analytics for proactive animal health management.
BeyondMath provides a generative physics platform for physical world simulation, built on the market’s largest AI model for real-time industrial scale engineering. The company is leveraging generative AI for live aerodynamic optimization and sustainable vehicle design. Through the Fellowship, BeyondMath is developing a generative AI foundation model for vehicle design that will further improve efficiency and sustainability.
Climate X delivers AI-powered risk analytics to banks, lenders, and asset managers, helping them build resilient real estate portfolios. Its flagship tool, Spectra, creates high-resolution digital twins for properties to assess and adapt to climate risks. As part of the Fellowship, Climate X is enhancing Spectra’s capabilities for simulating localized climate events and adaptation strategies.
Climavision is advancing global weather forecasting by addressing gaps in traditional methods, with a focus on energy, agriculture, and insurance sectors. As part of the Fellowship, Climavision is expanding its ClimaHD platform to power sub-kilometer weather predictions, leveraging HPC for improved accuracy.
One of the first two Australian companies to join the Fellowship, CounterCurrent reduces maritime emissions with AI-driven ship routing. Its platform combines generative ocean forecasting with reinforcement learning to optimize routes. The company is now working with the Fellowship to improve Hydra and Odysseus—its proprietary tools for real-time ocean prediction and adaptive route planning. By building its own generative AI foundation model, the team aims to help vessels cut fuel use by up to 25 percent without the need to change hardware.
Cusp AI is accelerating the discovery of breakthrough carbon capture materials by combining generative AI and computational chemistry. Having been accepted into the Compute for Climate Fellowship, the company is building CloudMiner, an AI-powered system capable of accelerating the design of metal-organic frameworks (MOFs) for climate applications. As part of that effort, the Cusp AI team are training a custom generative AI foundation model.
Eavor delivers next-generation geothermal energy through its closed-loop Eavor-Loop™ system. Eavor-Loop™ can operate in any geology without fracking and is being used to develop scalable and clean baseload energy solutions. To that end, the company is currently developing a POC for a modular, cloud-native multiphysics solver on AWS that uses AI to optimize drilling, system design, and autonomous geothermal power globally.
FortyGuard has built the world’s first AI-powered temperature operating system for cities, delivering real-time heat mapping and predictive cooling capable of reducing energy use by 30 percent. As part of the Fellowship, FortyGuard is building a new generative AI foundation model that combines satellite imagery and IoT data to help cities reduce energy consumption, protect public health, and build more climate-resilient infrastructure.
Fuse Energy is a UK-based full-stack renewable energy company that owns and operates solar and wind farms, integrating asset development with retail for direct consumer supply. As part of the Fellowship, Fuse Energy is developing an energy trading and risk management (ETRM) system for real-time monitoring and automated trading of renewable energy generation.
Matnex accelerates the discovery of sustainable materials by integrating quantum mechanics, AI, and high-throughput screening. The company is focused on sectors like batteries, automotive, and steel. In the Fellowship, Matnex is advancing SCOUT, a structure-based ML platform capable of screening over 10¹⁰ material compositions. A key part of that process involves developing a custom generative AI foundation model, that the company hopes will help them rapidly identify material compositions optimized for sustainability.
Muir AI’s product intelligence platform helps businesses reduce emissions and optimize costs. The company uses AI to model product emissions in minutes versus months across pharmaceutical and semiconductor industries, including Amazon. Muir AI is currently building Omni-BoM, an AI system that streamlines product emissions and cost analysis through universal Bill of Materials processing.
OnDeck AI automates visual monitoring of marine environments using vision-language models and retrieval-augmented generation (RAG). The company’s novel approach eliminates the need for labeled training data. Through the Fellowship, OnDeck is building a motion-grounded video reasoning system on AWS that can interpret unstructured marine video data in real-time to support conservation.
Pretred transforms waste tires into eco-friendly industrial barriers used in construction and infrastructure. As part of the Fellowship, the company is building SaferCrash SIM—a machine learning tool that predicts material performance and optimizes barrier design, helping reduce road fatalities and enable circular reuse of rubber waste.
Rainstick is an Australia-based startup and the first Indigenous-led business to join the Fellowship. The company combines AI with indigenous Aboriginal knowledge systems to develop bioelectric seed treatments for regenerative agriculture and ecosystem restoration. Rainstick is now creating "bioelectric seed recipes" that enhance crop yields, soil health, and the resilience of native habitats.
SXD is an AI design platform that generates zero waste product designs, helping to reduce CO2 emissions by 80 percent, material consumption by 69 percent, and costs by 55 percent. The company is now working with the Fellowship to develop To Zero by SXD AI, an algorithm capable of converting traditional product designs into zero waste versions by leveraging computer vision, machine learning, and combinatorial mathematics.
Windfall Bio develops novel solutions that use methane-eating microbes to convert methane into useful products like fertilizer. As part of the Fellowship, Windfall is developing an AI system that uses metagenomic data analysis to predict optimal microbial environments.
Xcimer Energy is pioneering a new path to clean baseload power through inertial fusion using excimer laser technology. As part of the Fellowship, the company is developing Iris, a digital twin platform built to optimize laser performance to accelerate the development and deployment of fusion energy systems.
Delivering the results that climate action demands
Each member of the Compute for Climate Fellowship’s 2025 cohort is committed to delivering outcomes to make our industries more sustainable, our businesses more efficient, and our planet more safe. Stay tuned to keep track of their progress in the months ahead.
Lisbeth Kaufman
Lisbeth Kaufman is the Founder and Head of the Climate Tech Startups BD team at Amazon Web Services. Her mission is to help the best Climate Tech startups succeed and reverse the global climate crisis through access to AWS’ cloud technology. Her team has technical resources, go to market support, and non-dilutive funding to help climate tech startups overcome obstacles and scale. With expertise at the intersection of climate and startups, Lisbeth was Founder and CEO of KitSplit.com, a sharing economy company called “the Airbnb of Cameras” by Forbes, and LucidHome.co, easy-to-understand climate risk reports for any address in the U.S. Before she was a founder, Lisbeth worked on climate policy as an energy/environment/agriculture policy advisor in the U.S. Senate. There she built a first-of-its-kind energy efficiency retrofit program and wrote a clean energy bill for farmers that got passed into law. Lisbeth has a BA from Yale and an MBA from NYU Stern where she was a Dean's Scholar. As a mentor at Techstars, Venture for Climate, and the Entrepreneurs Roundtable Accelerator Lisbeth mentors climate tech founders on product, growth, fundraising, as well as making strategic connections to teams at AWS and Amazon.
How was this content?