CodeFest 2026 on patent and IP portfolio (e)valuation
For the fourth edition of EPO CodeFest, we are excited to explore how code-based solutions can be used to assess the value of patents and intellectual property (IP) portfolios.
Making IP evaluation easier and more scalable can help startups and innovators access funding, bring new ideas to market and make smarter business decisions. By leveraging patent data, the economic value of IP can be unlocked, supporting growth, competitiveness and investment. These objectives are high on the agenda for European policymakers, as reflected in the EU Startup and Scaleup Strategy published on 28 May 2025. Join us as we push the boundaries of what's possible with patent data!
- Winners
58 people from 12 countries, organised into 15 teams and 9 individuals, submitted 24 proposals for the CodeFest 2026 competition.
For the first time in the competition, the jury was composed of both EPO staff and two external representatives, bringing the valuable perspective of future users into the evaluation process. The jury combined expertise in IT, data science, sustainability, AI, and patent information and analysis. The code solutions were evaluated by the jury using a holistic approach to assess the creativity and originality of systems designed to measure the economic and technological value of patents.
After two rounds of evaluation, the jury shortlisted six finalists, and the winners were announced during the CodeFest prize ceremony on 16 September, held as part of the PATLIB conference.
EPO CodeFest winners
Winners: Arnab Saha, Ralph Ryan Hebrio and Mauricio Rodriguez Alas
The Confused Electrons team developed an AI and machine learning powered platform that analyses patent portfolios and identifies the factors that contribute to a patent's long-term strategic value. By examining elements such as technological impact, market relevance and legal strength, the solution helps businesses understand how patents can support growth and create a competitive edge. The team received the grand prize of EUR 20 000.
- Arnab Saha is a doctoral engineer and CIFRE PhD researcher focusing on advanced RFIC design and semiconductor research and development
- Ralph Ryan Hebrio is a PhD researcher at 3D PLUS, focusing on resilient computational imaging systems, hardware characterization, and radiation testing environments.
- Mauricio Rodriguez Alas is a PhD researcher, co-founder of GoAlas.io, and CTO at Orbital Space Technologies.
First runner-up: George Abi Younes, Mehdi El Bouari, Reza Hosseini, Elena Mas Tur and Gaétan de Rassenfosse
Red-Cube coders created a solution that assesses both the technical strength and commercial value of patents. The solution introduces two key indicators: a "blocking score", measuring the extent to which patents may constrain future developments, and a "commercialisation score", estimating their market potential. Using machine learning and patent analytics, the solution analyses relationships between patents and indicators of technological and economic relevance. As first runner-up, the team received EUR 10 000.
- Gaétan de Rassenfosse leads the Science, Technology and Innovation Policy (STIP) Lab at EPFL, a leading centre in patent analytics.
- Reza Hosseini is the founder and CEO of Marsys.ai, a multi-agent artificial intelligence company.
- George Abi Younes is an Artificial Intelligence Manager at Artefact.
- Mehdi El Bouari is a recent EPFL graduate in Data Science.
- Elena Mas Tur is a European Patent Office (EPO) patent examiner and former faculty member at Eindhoven University of Technology (TU Eindhoven).
Second runner-up: Grazia Sveva Ascione and Nicolo Tamagnone
The team EUREKA developed an AI-driven system that mirrors the patent examination process by focusing on the inventive contribution of an invention in relation to the prior art. Rather than relying on simple statistical indicators, the platform analyses technical differences and identifies the closest prior art to assess the strength of an invention. The solution uses advanced AI models trained on patent examination data to go beyond basic keyword matching. By modelling examiner-style reasoning, it offers a deeper, patent-focused approach to evaluating innovation. As second runner-up, the team received EUR 5 000.
- Grazia Sveva Ascione is an assistant professor in economics and data science at the Polytechnic University of Turin. Her research includes entity linking for patent assignees, similarity models for patents, automatic classification of assignee types and the economics of patents.
- Nicolo Tamagnone is a senior machine learning engineer at Data Friendly Space (DFS), developing ML and NLP systems for the humanitarian and development sector.
- Finalists
Out of the 23 CodeFest 2026 proposals which advanced to the second round, we are proud to announce the 6 finalists selected by our jury of experts.
The finalists are presented in alphabetical order and this list does not reflect their ranking or placement.
Team Confused Electrons
For this year's challenge, the team developed an AI- and machine learning-based patent intelligence platform to address a key question: what makes a patent strategically valuable over time?
By assessing a combination of factors such as technological impact, market relevance and legal strength, the platform helps evaluate patent portfolios and identify long-term competitive advantages. A distinctive feature is its ability to link patent analysis with real-world business and technology trends.
Team EUREKA
The team developed an AI-driven system that mirrors the patent examination process by identifying the closest prior art to assess an invention’s strength. The system uses advanced AI models trained on patent data to go beyond basic keyword matching. By modelling examiner-style reasoning, this solution offers a particularly sophisticated approach to evaluating innovation.
Lucas Prieels
Lucas developed an AI-assisted extension of the EPO’s IPscore methodology, combining a well-established evaluation framework with modern machine learning to enable faster and more data-driven patent portfolio analysis. The system uses AI and natural language processing to assess factors such as citation impact, renewal likelihood and technological relevance, while supporting users in completing qualitative evaluations. Results are integrated into a familiar IPscore-style workflow, making them easy to interpret and apply.
Marc Donovici
Marc developed a machine learning framework that uses citation data, semantic analysis and predictive models to assesses the long-term, economic value of patents based on analysis of their technological influence, legal robustness and commercial potential. The system takes into consideration aspects such as patent renewal patterns and citation impact to identify patents with sustained value. Designed with a strong focus on transparency, the framework provides clear insights into the factors driving patent value, supporting more informed IP management and investment decisions.
Team Red-Cube-Coders
The team developed an AI and machine learning-based approach to assess the technical strength and commercial value of patents. The solution introduces two key indicators: a “commercialisation score”, estimating their market potential, and a “blocking score”, measuring how patents may constrain future developments. Using machine learning and patent analytics, it analyses relationships between patents and signals of technological and economic relevance. By focusing on real-world impact, the framework supports a more business-oriented understanding of patent portfolio value.
Team Turbo KD
The team developed the IP Impact Index, an AI-assisted framework designed to translate complex patent data into clear, business-oriented insights. The platform uses a multi-dimensional approach to assess patents based on factors such as technological significance, market relevance and strategic positioning, providing transparent and explainable results rather than a single opaque score. By combining patent data with structured analytics and intuitive presentation, the solution helps stakeholders better understand and use patent information for decision-making.
Join us for the online prize ceremony on 16 September
Eager to see who will take the top spot? Register for our PATLIB Conference to attend the online CodeFest 2026 prize ceremony on 16 September.
- The coding challenge
Create an automated system that uses data analytics to assess patent value.
Potential approaches to evaluating IP value could look like but are not limited to:
- Technological importance: determine how unique, disruptive or broadly applicable a technology is
- Market-specific valuation: assess the technology’s market readiness, the importance of its potential market and existing competition from alternative technologies
- Legal strength: estimate the likelihood a patent will be granted, stand up to legal challenges or be successfully enforced
Participants can focus on a single topic (e.g. NLP-based technology breadth, monetary valuation or predictive legal models) or combinations thereof. Mixed teams comprising AI specialists, economists and legal and IP experts are encouraged, as multidisciplinary collaboration can help ensure impact.
This challenge offers a unique opportunity to contribute to the EPO’s efforts to maximise the power of patent data and enhance the evaluation of IP.
Want to participate?
This challenge is of particular interest to patent professionals, IT and AI specialists, economists, legal and IP experts, data enthusiasts, developers, coders, data scientists and professionals from international organisations, academia and research institutions. Only residents of EPO member states aged 18 and above can participate aged 18 and above can participate. Submit your proposals for solutions by 27 November 2025, 23.59 CET. Submissions can come from an individual person or a team of up to five people. Teams made up of EPO staff members, external entrants or a mixture of both are all welcome! Submit your proposals for solutions by 27 November 2025, 23.59 CET. Submissions can come from an individual person or a team of up to five people. Teams made up of EPO staff members, external entrants or a mixture of both are all welcome!
In line with the EPO's efforts to promote diversity and inclusion across the innovation landscape, women with an enthusiasm for technology and innovation are highly encouraged to take part, showcase their skills and help inspire the next generation.
In order to participate you must meet all eligibility criteria included in the rules of competition at the time of submitting your application.
Steps
- Submit your proposal by 27 November 2025, 23.59 CET. It should include:
- The completed form with your personal information. Please note that all team members must register individually and indicate if they are part of a team.
- An outline of your proposal for solving the coding challenge. Proposals must include an explanation of your suggested approach and indicate the data selection you intend to use, possible model implementation and model evaluation.
- During the first selection process, the submitted solutions are evaluated based on the rules of competition to decide which will move on to the next round.
- At the beginning of December 2025, participants will be informed if their proposal has made it to the next round.
- For the second round, selected participants will work on executing their proposed solution and submit their code by 28 April 2026 at 23.59 hrs CET.
- The jury will evaluate the final solutions based on criteria such as creativity, functionality and technical implementation.
- Finalists will be announced in June/July 2026 and the winners will be revealed at the prize ceremony on 16 September 2026.
Prizes
The winners of Code Fest 2026 will be eligible to win one of the following prizes:
- Grand prize winner: EUR 20 000
- First runner-up: EUR 10 000
- Second runner-up: EUR 5 000
Each prize will be accompanied by a trophy.
- Jury
Submissions will be evaluated by senior specialists from across the EPO working in IT, data science, artificial intelligence, sustainability and patent information and analysis.
Yann Ménière, EPO Chief Economist, Patent Research and Policies
Yann Ménière joined the EPO as Chief Economist in 2016. He leads a team responsible for the preparation of economic studies, patent analytics and digital tools within the EPO Observatory on Patents and Technology. Yann was previously a professor of economics at Université Paris Sciences et Lettres where he led the chair on IP and Markets for Technology. He has many years of academic and field experience in issues relating to the economics of patents, innovation, technological change and economic growth.
David Horat, Director of Certifications and Educational Technologies, Directorate-General 5 Legal and International Affairs
David Horat oversees the European Qualifying Examination (EQE), the European Patent Administration Certification (EPAC) and the infrastructure of the European Patent Academy. Previously he was Director of Patent Knowledge where he led the development and delivery of data, information and knowledge products and services. He began at the EPO as a patent examiner and was key in the development of the first AI pre-search algorithm. He later headed the Prior-art Data Management department, leading a team of over 150 people across 10+ locations, delivering 1 000+ data pipelines, 2+ petabytes of data and introducing new innovative tools. Before joining the EPO, David worked as a software engineer at the European Organization for Nuclear Research (CERN) and Ericsson. David holds an MSc in computer science, an LLM in intellectual property law and business management, and he has completed an executive programme in leadership.
Nathalie Sechi, Team manager of Business Analysis, Business Information and Technology, EPO
Nathalie Sechi joined the EPO in 2015 and has 25 years of experience in business architecture and business and strategy analysis. She has contributed extensively to business and digital transformation initiatives, with a clear focus on delivering tangible business value in everything from strategy to change implementation. Her strength lies in identifying business challenges that can be overcome using IT. Nathalie holds a master's in industrial economics and a PhD in industrial automation, with a focus on decision-making in the innovation process.
Volker Hähnke, Data Scientist, Business Information and Technology, EPO
Volker Hähnke is a data scientist with a background in bioinformatics and cheminformatics who studied at Goethe University Frankfurt and ETH Zürich. After a postdoctoral fellowship at the U.S. National Center for Biotechnology Information, where he contributed to PubChem’s chemical structure standardisation, he joined the EPO in 2015. In his role, he applies machine learning and natural language processing to improve patent search and prior art retrieval, with a focus on learning-to-rank models, user behaviour analysis and search evaluation. He brings deep technical expertise and a passion for applied AI to everything he does.
Hélène Dernis, Statistician and Analyst, Directorate for Science, Technology and Innovation (STI), OECD
Hélène Dernis is responsible for the STI Micro-data Lab, a data infrastructure that collects and links large-scale administrative data, including IP filings and commercial micro-level datasets, and aims to provide measures related to innovation and entrepreneurship. Since joining the OECD, she has contributed to the development of IP-related indicators, notably on patents, to support evidence-based policy.
Julie Callaert, Research Expert, Research and Development Monitoring Research Centre ECOOM, KU Leuven, Belgium
Julie Callaert manages the centre's technometrics study group, which provides patent indicators to the Flemish government to support their technology policy-making. Her research is primarily oriented towards the development of novel and enhanced patent indicators, with a particular interest in the dynamics of science and technology interplay, including academic entrepreneurship. She has published on these topics in journals like Research Policy; Scientometrics; Journal of Technology Transfer; Economics of Innovation and New Technology; and Industry and Innovation. With her background, and as a member of the INCENTIM research division from the KU Leuven’s Faculty of Business and Economics, she has also been involved in several European Framework projects.
- Timeline
The competition has five key stages:
16 September 2025 to 27 November 2025
Entrants must register and submit their proposals by 27 November 2025, 23.59 hrs CET. Proposals must include an outline and explanation of the suggested approach as well as indicate the intended data selection, the possible model implementation and model evaluation.
28 November to 5 December 2025
The proposals will be evaluated based on the rules of competition to decide which will move on to the next round.
5 December 2025 to 18 April 2026
The coding challenge is on!
All successful candidates will be given access to the EPO's data resources, including datasets and services. For the competition, the EPO will provide participants with a special CodeFest profile that includes enhanced computational resources in the Technology Intelligence Platform.
Access will be given to:
- The Technology Intelligence Platform, the EPO’s next-generation tool for processing, analysing and visualising patent data. Offering a JupyterLab based environment, it features a powerful computing infrastructure and provides easy access to the patent data contained in PATSTAT Global, PATSTAT EP Register and EP full-text through EPO-developed libraries.
- EPO worldwide bibliographic data (DOCDB)
- EPO worldwide legal event data (INPADOC)
- EP full-text data
- Open Patent Services
Consider that not all data resources will be needed to implement your solution. To submit a strong proposal based on the available data sources, look at which bulk data sets best suit your needs.
The coding challenge will run for over twelve weeks and code must be submitted by 28 April 2026. at 23.59 hrs CET.
June/July 2026
Announcement of the finalists!
Finalists are invited to the EPO CodeFest 2026 prize ceremony.
16 September 2026
EPO CodeFest 2026 prize ceremony!
During the ceremony, the winners will be announced. Details of the full programme to follow.
- Registration
Registration closed.
Registration deadline
27 November 2025, 23.59 hrs CET.Data Protection Statement (DPS)
DPS for CodeFest 2026Rules of Competition (RoC)
CodeFest 2026– Rules of competitionBackground
Patent valuation is playing an increasingly critical role in driving business success across all technology fields. Under its Strategic Plan 2028, the EPO is advancing methods for IP valuation to help startups and inventors secure funding and scale sustainably within Europe’s innovation ecosystem. By leveraging patent data to streamline IP evaluation, innovators can bring new ideas to market more effectively and make better-informed, strategic business decisions. These goals align closely with current policy priorities, as highlighted in the EU Startup and Scaleup Strategy published on 28 May 2025. This CodeFest challenge contributes to the EPO’s efforts to develop advanced tools to help innovators assess the value of their patent portfolios to make smarter IP and investment decisions.
Further information
CodeFest 2023 on green plastics
CodeFest 2024 on generative AI
CodeFest Spring 2025 on classifying patent data for sustainable development
