
CleaRIS
Cleantech Recommendation and Intelligence System

Abstract
The global climate goals are ambitious. Together with tightening sustainability regulations such as the Corporate Sustainability Reporting Directive (CSRD), they put tremendous pressure on companies to get more innovative and leverage new technological developments in CleanTech and ClimateTech to their maximum. The International Energy Agency estimates that half of the greenhouse gas emissions that are to be reduced by 2050 can only be achieved through new technologies that don't yet exist. On the other hand, countries and businesses struggle to get up to speed with climate innovation, as evidenced by more and more cases of climate-related fines (e.g., as imposed by the ECB in May 2024) and international court litigations (e.g., the Swiss “climate seniors” case in April 2024). While it is clear that businesses need to take action, in most cases, they lack the guidance and expertise needed for efficient climate innovation. In this project, we aim to build an intelligence platform that continuously monitors and visualizes the global CleanTech industry. To make these insights actionable, it provides tailored recommendations to businesses about implementing CleanTech innovations, fulfilling regulatory requirements, and communicating CleanTech initiatives to different stakeholders (consumers, investors, etc.). The platform combines traditional predictive AI methods with new developments in generative AI, which allows us to reduce hallucinations and ground recommendations and predictions in a solid database of market insights. These insights are extracted from a large-scale database containing a variety of public sources which are analyzed using predictive AI methods. Recommendations are generated using an open-source Large Language Model (LLM), such as Llama 3.2, which will be fine-tuned to the CleanTech domain. Using Retrieval-augmented Generation (RAG), the system not only has access to a wide range of public CleanTech sources, but can also access company-internal documents of individual clients, thus adapting the insights and recommendations to their unique context and needs.
In this collaboration we aim to pair SDSC’s capacity of delivering innovative, reliable and robust AI products with the collaborators’ broad academic track record in the domains such as ESG factors, CleanTech innovation, greenwashing, and the AI-driven decision-making as well as the extensive experience of delivering LLM-based models for education and research, in order to build this platform that can accelerate the global transition to green economy.
People
Collaborators


Anna joined SDSC as a Data Scientist focusing on industry collaborations in July 2019. She completed her PhD in Bioinformatics at the University of Luxembourg, where she analysed large-scale heterogeneous datasets and leveraged multiple disciplines: Statistics, Network Analysis, and Machine Learning. Before joining SDSC, Anna worked as a Data Scientist at Deloitte Luxembourg, with a focus on computer vision and time-series analysis.Currently, Anna is a Principal Data Scientist based at the ETH Zurich office, where she leads biomedical collaborations with industry partners. Anna works on a range of projects: protein properties prediction, biomanufacturing optimization, statistical model evaluation and others.


Snežana joined the SDSC industry team in June 2021 on a mission to advance adoption of modern data driven solutions in the domain of public health care. She has a background in experimental particle physics with a Diploma from the ETH Zurich and a PhD from the University of Geneva. Snežana pursued fundamental research in the field of high energy physics at CERN for nine years, harnessing the power of machine learning and statistical methods to uncover the traces of new physics in petabytes of proton-proton collision data and to develop innovative particle identification algorithms. Since 2018, Snežana served as a Data Science consultant, supporting partners from industries such as manufacturing, insurances, compliance services and online platforms in creating business value from internal and external data.
description
Motivation
Currently, a group of researchers from ETH, HSLU, and FHNW are conducting projects focusing on Environmental, Social, and Governance (ESG) factors, Cleantech innovation, greenwashing, and the integration of AI into decision-making processes for SMEs. The selected industry partner, Anacode, develops AI systems for customers in industries such as aerospace, automotive and commodities, where Cleantech innovations play a major role. Our collaborators have extensive experience of delivering LLM-based models for education and research events. They would like to build a flagship collaboration amongst the institutions for this strand of research.
Proposed Approach / Solution
Our collaborators aim to build an intelligence platform that continuously monitors and visualizes the global Cleantech industry. It provides tailored recommendations to businesses about implementing Cleantech innovations, fulfilling regulatory requirements, and communicating Cleantech initiatives to different stakeholders. In order to have first results for further analysis and improvement and to obtain a fast-to-prototype product, the project will start with a focus on the automotive and aerospace industries.
The core objectives include:
- Building a Cleantech-specific AI platform that provides recommendations and guidance for Cleantech adoption and innovation that are aligned with relevant regulations.
- Fine-tuning an LLM such as DeepSeek for domain-specific knowledge integration.
- Validating the Retrieval-Augmented Generation (RAG) system, ensuring precise, fact-based AI responses.
- Exploring various methods for embedding Cleantech expertise into the conversational system, including knowledge graphs, structured prompts, and neuro-symbolic AI (as described in the article series Injecting domain knowledge into your AI system).
Presentation
Gallery
Annexe
Additional resources
Bibliography
Publications
Related Pages
- Renku project: Reproducible Data Science | Open Research | Renku
More projects
WHO-Hive
MAGNIFY
News
Latest news


PAIRED-HYDRO | Increasing the Lifespan of Hydropower Turbines with Machine Learning
PAIRED-HYDRO | Increasing the Lifespan of Hydropower Turbines with Machine Learning


First National Calls: 50 selected projects to start in 2025
First National Calls: 50 selected projects to start in 2025


AIXD | Generative AI toolbox for architects and engineers
AIXD | Generative AI toolbox for architects and engineers
Contact us
Let’s talk Data Science
Do you need our services or expertise?
Contact us for your next Data Science project!