Announcement, Company

Pruna AI secures $6.5 million to sustainably optimize machine learning models and empower AI engineers to focus on innovation

Nov 18, 2024

Bertrand Charpentier

Founder, President & Chief Scientist

Pruna AI secures $6.5 million to sustainably optimize machine learning, models and empower AI engineers to focus on innovation

  • Pruna AI is the AI optimisation engine to make any AI model faster, cheaper, smaller, and greener.


  • The company’s optimization engine combines a variety of techniques to rapidly adapt and enhance models without the need for extensive reengineering, delivering superior compression results in a snap.


  • With Pruna AI, engineers can focus their time, effort, and budget on product delivery rather than painstakingly engineering, compressing and finetuning models for every deployment.


  • Pruna AI’s solution also addresses major environmental concerns surrounding AI, with compressed models reducing AI-related carbon emissions by up to 91%.


Top: from left to right, Rayan Nait Mazi, Co-Founder & CEO and Bertrand Charpentier, Co-Founder & Chief Scientist

Bottom: from left to right, John Rachwan, Co-Founder & CTO, and Stephan Günnemann, Co-Founder & Chief Strategist


Munich and Paris, 18 November, 2024. Pruna AI, the AI optimization engine empowering ML teams to simplify code and streamline models, today announces that it has successfully raised $6.5 million in a seed funding round led by EQT Ventures with participation from Daphni, Motier, and Kima, along with support from high profile angels including Roxane Varza, Hervé Nivon, and Olivier Pomel. The company will use the fresh capital to expand its technical team to a full staff of machine learning research engineers, and operation and go-to-market specialists.

The exponential growth of AI is unprecedented. Bain estimates that the total addressable market for AI-related hardware and software will grow between 40% and 55% annually for at least the next three years, reaching between $780 billion and $990 billion by 2027. In the midst of this global AI arms race, companies of all kinds are sprinting to keep pace with the rapid developments taking place, leading to squeezed resources and skyrocketing costs. Research from Epoch AI suggests that the cost of training and developing models has almost tripled since 2016 - for businesses with smaller budgets, this presents a very real barrier to innovation.

Developed by leading ML researchers from top institutions including the Technical University of Munich, and with over 30 years of combined experience and a portfolio of 270+ published papers, Pruna AI’s founders have profound industry and technological expertise. The team recognized the need for AI developers to find more efficient ways to tailor models to meet the rapidly changing needs of businesses and developed Pruna AI - the AI optimization engine for ML teams looking to streamline their work and cut costs. Using compression methods including pruning, quantization, compilation and caching, among others, the company’s platform enables AI developers to simplify deep learning models in a snap, making them more efficient across hardware setups.

Designed to be used by companies of all sizes, Pruna AI enables AI engineers without Big Tech levels of resources at their fingertips to focus on their R&D efforts - taking care of the painstaking optimization process so they don’t have to. The solution supports all major compression methods, making it a seamless addition to any machine learning pipeline.

Beyond time and cost savings, Pruna AI presents a path to sustainable AI. While AI continues to revolutionize a wide range of industries at lightning speed, models consume a significant amount of energy when deployed at scale. On average, a ChatGPT query needs nearly 10 times as much electricity to process as a Google search

With companies under increasing pressure to improve their sustainability performances, reducing their AI-related carbon footprints is a top concern. Pruna AI addresses this by compressing machine learning models, reducing the energy and computational power required to run them, and allowing engineers to cut carbon emissions by up to 91%. Additionally, Pruna AI compressed models reduce the strain on hardware, enabling it to last longer without the need for frequent replacements." In a tech landscape where AI advances show no sign of slowing, Pruna delivers a sustainable, efficient, and accessible solution to support everyday developers to leverage the performance and scaling benefits of AI at significantly lower financial and environmental costs.

Charpentier, Co-Founder, President, and Chief Scientist at Pruna, commented:

“EQT Ventures’ investment in Pruna AI comes at a pivotal time in our business’ growth and we are looking forward to scaling and expediting our product launch with their support. The proceeds from our raise will also go towards helping to grow our talented team – which has already tripled in size – enabling us to fulfill our goal of helping to support engineers develop AI in the fastest, cheapest, and most sustainable way possible. EQT Ventures were a clear choice of partner – they are experts in supporting young, ambitious businesses to scale and innovate sustainably for the future, which is at the heart of what we do."

Carl Svantesson, Partner at EQT Ventures, added:

“Striking the right balance between growth and environmental accountability is top of mind for businesses of all sizes these days. By bridging efficiency and sustainability, Pruna AI not only democratizes cutting-edge technology but also champions a future in which every company can scale rapidly and responsibly. The team’s Munich-Paris ecosystem affords them extra leverage in a competitive market, giving them a strong commercial foothold and a compelling talent acquisition opportunity. We are thrilled to welcome Pruna AI to the portfolio.”


About Pruna AI

Pruna AI is the AI optimization engine for ML teams looking to simplify their work. It makes running AI models cheaper, faster, and greener, empowering ML engineers to focus on innovation. The company’s solution allows users to optimize and run deep learning models in a snap across any hardware setup. It can be integrated directly onto companies’ existing AI systems, leaving their existing infrastructure untouched and providing immediate benefits without additional overhead. Created by a team of leading ML efficiency and reliability researchers, Pruna AI’s platform has been proven to cut carbon emissions by up to 91% compressing models to deliver results quicker, reduce time-to-market, and help businesses reach their sustainability goals.

About EQT Ventures

EQT Ventures is an early stage lead investor built by founders and operators, offering the next generation of entrepreneurs a fast track to scale. The fund is based in Luxembourg and has investment advisors strategically located across Stockholm, Amsterdam, London, New York, Berlin, and Paris. Currently investing out of its third (€1.1B) fund, the largest early-stage fund ever raised in Europe, EQT Ventures is one of the most active VC firms partnering with hundreds of ambitious founders and startups. Driven by a team of accomplished company builders and scalers, EQTVentures is committed to providing the capital and hands-on support necessary for generation-defining founders and companies to transform their visions into global successes.

Press contacts:

  • Pruna AI spokesperson: Bertrand Charpentier – bertrand.charpentier@pruna.ai

  • EQT Ventures contact: Paul Andrieu, Kekst CNC – paul.andrieu@kekstcnc.com +44 7581 053448

Button

Button

Bertrand Charpentier

Nov 18, 2024

Bertrand Charpentier

Nov 18, 2024

Bertrand Charpentier

Nov 18, 2024

Speed Up Your Models With Pruna

Inefficient models drive up costs, slow down your productivity and increase carbon emissions. Make your AI more accessible and sustainable with Pruna.

Speed Up Your Models With Pruna

Inefficient models drive up costs, slow down your productivity and increase carbon emissions. Make your AI more accessible and sustainable with Pruna.

Speed Up Your Models With Pruna

Inefficient models drive up costs, slow down your productivity and increase carbon emissions. Make your AI more accessible and sustainable with Pruna.

© 2024 Pruna AI - Built with Pretzels & Croissants 🥨 🥐

© 2024 Pruna AI - Built with Pretzels & Croissants

© 2024 Pruna AI - Built with Pretzels & Croissants 🥨 🥐