New network announcement: the Swiss Personalized Health Network (SPHN) launches a new cross-hospital data query tool powered by Tune Insight! Swiss Personalized Health Network (SPHN) is introducing the Data Exploration and Analysis System (DEAS), poised to transform data-driven research in Switzerland. It offers researchers improved tools for conducting data queries and feasibility studies. This platform will replace the current Federated Query System, which will be discontinued by December 2024. DEAS, developed in collaboration with SPHN and Swiss university hospitals, will enable secure access to aggregated data from Swiss hospitals. Powered by Tune Insight, DEAS complies with the highest levels of data protection, based on state-of-the-art privacy technologies including homomorphic encryption, secure multiparty computation, federated analytics and differential privacy. For more details, read the announcement on SPHN’s website: https://lnkd.in/e9QHs9Cj Team Tune Insight: Juan Ramón Troncoso-Pastoriza, Romain Bouyé, Frederic Pont, Jean-Pierre Hubaux, Manon M., Yassine Boukhari, Florimond Houssiau, Kadiatu Jalloh, MSC, Thierry Bossy, Taisiia Efremova, Boris Flesch, Adrien Prost, Loïs Huguenin-Dumittan, Mariya Georgieva, Francesco Marino, Bertrand Spilthooren #personalizedhealth #federatedanalytics #datacollaboration #dataprotection #dataprivacy #federatedlearning #swissstartups
Tune Insight
Software für Datensicherheit
Encrypted computing platform for data collaborations, valorization and machine learning
Info
Tune Insight is a Swiss B2B startup at the forefront of cutting-edge technology with an unwavering commitment to data security and privacy. Our Encrypted Computing solution offers unparalleled security and risk minimization by allowing computations to be performed in a federated manner on encrypted data so that your sensitive information never leaves your organization perimeter and remains confidential throughout the process. This allows for secure data collaborations, valorization and monetization. Our mission is to transform the paradigm of the data economy into an insight economy that better protects sensitive data, that is more secure, fair, and protective of privacy and confidentiality rights
- Website
-
https://tuneinsight.com
Externer Link zu Tune Insight
- Branche
- Software für Datensicherheit
- Größe
- 11–50 Beschäftigte
- Hauptsitz
- Lausanne
- Art
- Privatunternehmen
- Spezialgebiete
- data privacy, data valorization, data protection, homomorphic encryption, multiparty computation, federated learning, differential privacy, synthetic data, privacy-enhancing technologies (PETs), generative AI, LLMs, healthcare und financial services
Orte
-
Primär
EPFL
Lausanne, CH
Beschäftigte von Tune Insight
-
Jean-Pierre Hubaux
EPFL Professor, co-founder of Tune Insight
-
Pascal Mathis
Partner @Founderful, Co-founder @GetYourGuide
-
Bertrand SPILTHOOREN
Accélérateur de l’Innovation en Santé | Solutions Fédérées | Utilisation secondaire sécurisée des données de santé
-
Juan Ramón Troncoso-Pastoriza
Co-Founder and CEO at Tune Insight
Updates
-
🚀 A Major Milestone for Secure and Federated Genomic Research! We are thrilled to announce another significant scientific achievement for Tune Insight! In collaboration with Yale University, Broad Institute of MIT and Harvard, and EPFL, our latest research has just been published in Nature Genetics: "Secure and federated genome-wide association studies for biobank-scale datasets." 🔬 In this paper, we introduce Secure Federated Genome-Wide Association Studies (SF-GWAS)—a groundbreaking approach combining secure computation frameworks and distributed algorithms. SF-GWAS enables efficient and accurate genomic studies across private datasets held by multiple institutions while guaranteeing data confidentiality. 📊 Our work demonstrates unprecedented accuracy and efficiency, including a UK Biobank cohort of 410,000 individuals, achieving an order-of-magnitude improvement in runtime over previous methods. This marks a major step forward in secure, large-scale collaborative genomic research. We are excited about the possibilities this unlocks for the future of privacy-preserving healthcare innovation! Read the full article here: https://lnkd.in/exY8MHuK Authors: Hyunghoon Cho, David Froelicher, Jeffrey Chen, Manaswitha E., Apostolos Pyrgelis, Juan Ramón Troncoso-Pastoriza, Jean-Pierre Hubaux, Bonnie Berger #FederatedLearning #Genomics #PrivacyPreservingAI #HealthcareInnovation
-
🚨Improving LLM privacy while maintaining performance! Tune Insight announced this week at AMLD - Applied Machine Learning Days that we demonstrated how to reduce memorization in LLM fine-tuning with sensitive medical data up to a factor of 10, with our research partners EPFL and Yale University! - - 🚨Améliorer la confidentialité des LLMs tout en maintenant leurs performances! Tune Insight annonce cette semaine à AMLD - Applied Machine Learning Days que, en collaboration avec ses partenaires de recherche EPFL et l'Université de Yale aux Etats-Unis, nous avons démontré comment réduire jusqu'à un facteur de 10 la mémorisation des données médicales sensibles lors du fine-tuning de LLM! - - Title: Mitigating Unintended Memorization with LoRA in Federated Learning for LLMs Full paper at https://lnkd.in/dxDBfYXb Authors: - Thierry Bossy (Tune Insight) & Julien Vignoud (EPFL), equal contributions - Tahseen Rabbani (Yale University), Juan Ramón Troncoso-Pastoriza (Tune Insight), Martin Jaggi (EPFL) Abstract: Federated learning (FL) is a popular paradigm for collaborative training which avoids direct data exposure between clients. However, data privacy issues still remain: FL-trained large language models are capable of memorizing and completing phrases and sentences contained in training data when given with their prefixes. Thus, it is possible for adversarial and honest-but-curious clients to recover training data of other participants simply through targeted prompting. In this work, we demonstrate that a popular and simple fine-tuning strategy, low-rank adaptation (LoRA), reduces memorization during FL up to a factor of 10. We study this effect by performing a medical question-answering fine-tuning task and injecting multiple replicas of out-of-distribution sensitive sequences drawn from an external clinical dataset. We observe a reduction in memorization for a wide variety of Llama 2 and 3 models, and find that LoRA can reduce memorization in centralized learning as well. Furthermore, we show that LoRA can be combined with other privacy-preserving techniques such as gradient clipping and Gaussian noising, secure aggregation, and Goldfish loss to further improve record-level privacy while maintaining performance. Team Tune Insight: Juan Ramón Troncoso-Pastoriza, Romain Bouyé, Frederic Pont, Jean-Pierre Hubaux, Manon M., Yassine Boukhari, Florimond Houssiau, Kadiatu Jalloh, MSC, Thierry Bossy, Taisiia Efremova, Boris Flesch, Adrien Prost, Loïs Huguenin-Dumittan, Mariya Georgieva, Francesco Marino, Bertrand Spilthooren, Amal MANSOURI, Eya Chebbi #AMLDEPFL2025 #AI #MachineLearning #LLMs #GenAI #FederatedLearning #DataProtection #DataPrivacy
-
-
Tune Insight hat dies direkt geteilt
🎥 Talk with our Technical Lead Mariya Georgieva 📢 👉 Check out our latest interview with Mariya Georgieva, Technical Lead at Tune Insight! In this interview, Mariya shares insights into her contributions to PRIVIDEMA Horizon Europe project 101167964, the role of Tune Insight in the project, and the most exciting and impactful aspects of #PRIVIDEMA. She also discusses her expectations, the biggest challenges, and what excites her most about working on this innovative project. 👉 https://lnkd.in/d4wd7SHe European Cybersecurity Competence Centre (ECCC) & Cybersec Europe #HorizonEurope #EU #CyberSecurity #dep #PrivacyandIdentityManagement #biometricdata #authenticationfactor #identitydigitalwallet #cyberthreatintelligence #homomorphicencryption #federatedlearning
PRIVIDEMA - Mariya Georgieva (Technical Lead)
https://vimeo.com/
-
We will be at AMLD - Applied Machine Learning Days at EPFL in Lausanne🇨🇭 this week, to discuss the role of secure data access to train AI, and how our platform helps reconcile AI progress and data privacy. You can find us: - All week long on our booth, with our experts Thierry Bossy (ask him about measuring data leakage when fine-tuning LLMs on sensitive data, and data harmonization with L(ight)LMs ) and Ann-Sophie Saborowski (ask her about our hospitals networks and how we provide CROs and bio-pharmas accelerated access to insights they need from data they don't have) - At Swisscom’s session on cybersecurity and AI on Tuesday afternoon, organized by David Froelicher and Daniel Dobos, with our CEO Juan Ramón Troncoso-Pastoriza: Federated Health Data Space, Secure Data Exploration and Analysis from Feasibility to AI - At the startup pitch sessions on Tuesday morning and Thursday afternoon, with our CEO Juan Ramón Troncoso-Pastoriza At Tune Insight, we provide secure data analytics, machine learning and LLMs. By integrating federated learning, homomorphic encryption, differential privacy and synthetic data, we enable collaborative development and secure use of machine learning models and LLMs, ensuring compliance with data regulations and protecting sensitive information in sectors like healthcare, bio-pharma and finance. ✅ Secure: GDPR compliant, more secure than federated learning; data never transferred or revealed. ✅ Scalable: more scalable than traditional multi-party computation. Manage hundreds of participants. ✅ Collective: as more organizations join the collaboration, insights get more precise and valuable. ✅ Performant: highly parallelizable, we bring better performance than fully homomorphic encryption. ✅ Flexible: fully software-based, no need for trusted third party hardware or a new infrastructure. #AMLDEPFL2025 #AI #MachineLearning #LLMs #GenAI
-
-
🇫🇷 📹 Découvrez le portrait de Tune Insight par Bertrand Spilthooren, réalisé sur Les "Grandes Tendances de la E-santé" par L'Oeil de la E-santé! ✅ 🤝 nous permettons aux acteurs de la santé, aux établissements de soins, aux CROs et à l'industrie bio-pharmaceutique de collaborer en toute sécurité de manière fédérée ✅ 🔒 les données brutes ne quittent jamais leur périmètre actuel de sécurité ✅ 💡 les connaissances collectives (les "insights") sont extraites avec des garanties régulatoires, sans compromettre la souveraineté ni la confidentialité des données ✅ 📈 📊 🚀 de l'appariement aux études de faisabilité, des statistiques à la validation d'hypothèses jusqu'à l'entrainement d'algorithmes d'IA, notre plateforme supporte tout le spectre des sciences de la donnée, sans compromis sur la protection des données, au service de la santé #ESANTE25
#ESANTE25 - TUNE INSIGHT
https://www.youtube.com/
-
🗞️ "La solution de Tune Insight permet de traiter les données dans le secteur de la santé en tout sécurité, et utiliser l'IA en toute confiance!" Merci Marie de FREMINVILLE pour ce bel article dans InCyber News! https://lnkd.in/e9RgJkDj
Traiter les données dans le secteur de la santé en toute sécurité, et utiliser l’IA en toute confiance: :les solutions suisses ! - INCYBER NEWS
incyber.org
-
We are at EUCROF 2025 in Copenhagen 🇩🇰 until Tuesday, February 4th! Meet Kadiatu Jalloh, MSC and Bertrand Spilthooren to discuss how our federated health data space solution connects networks of hospitals to the bio-pharmaceutical industry to accelerate timelines, reduce costs and risks, and foster innovation for clinical research!
-
-
Tune Insight hat dies direkt geteilt
🔎 Partager des données sensibles sans les déplacer ni les divulguer ? C'est possible ! 📊 Dans le secteur de la #santé, des années peuvent s’écouler entre la définition d’une étude et l’accès aux données. Trop long, trop contraignant. 🤔 Comment garantir la #confidentialité tout en facilitant la recherche médicale ? 💡 Comment appliquer ces principes à la finance ou à l’#IA ? 🇨🇭 Une solution née à l'École polytechnique fédérale de Lausanne, EPFL et développée par Tune Insight permet d’exploiter les données de manière sécurisée grâce au #ZeroTrust. 👉 Grâce à Marie de FREMINVILLE, découvrez comment cette technologie révolutionne la gestion des données sensibles. https://lnkd.in/e9RgJkDj #cybersécurité #intelligenceartificielle #InCyberNews Forum InCyber Europe (FIC) ID FORUM
Traiter les données dans le secteur de la santé en toute sécurité, et utiliser l’IA en toute confiance: :les solutions suisses ! - INCYBER NEWS
incyber.org
-
Tune Insight hat dies direkt geteilt
Un grand #Merci à Interaction Healthcare pour ce très bel événement hier à STATION F : l’incontournable Les "Grandes Tendances de la E-santé" 2025, neuvième édition. La journée fut riche de rencontres, d’échanges et d’un programme complet de très haut niveau et nous remercions toutes les personnes nous ayant fait l’amitié de venir à notre rencontre pour parler d’accélération de la #recherche, de modèles #fédérés et #sécurisés pour une #IA #souveraine, #Responsable, #Ethique et animée par la volonté de contribuer au #bienCommun , au #respect des libertés et droits fondamentaux des #patients. Face à la vision desormais dérégulée des États Unis, à l’émergence de la Chine avec DeepSeek, Tune Insight propose la plateforme idéale pour permettre à l’Europe de rester compétitive à la pointe de l’innovation en recherche thérapeutique et IA sans compromis sur sa #doctrine respectueuse des individus à laquelle nous sommes si attachés. J’y ai même revu un très vieil ami de Sup de Co Marrakech que je n’avais pas vu depuis 15 ans, Morad El Mazyani qui venait porter son entreprise de formation aux nouvelles technologies d’IA dans un autre événement organisé dans les locaux et l’arrière petit neveu d’Antoine de Saint-Exupery, Raphaël de Saint-Exupery Avec l’équipe Tune Insight, j’ai été enchanté de participer car d’une part, cela me tenait à cœur et d’autre part, ce fut vraiment un moment exceptionnel dans un temple de l’innovation. #Merci à Monsieur le Minsitre Yannick NEUDER de nous avoir fait l’honneur de poser avec nous, à Hela Ghariani pour son intervention, David Sainati que j’ai croisé et avec qui j’ai toujours plaisir à discuter, au toujours aussi exceptionnel ♦️David PETAUTON♦️ dont l’amitié m’honore, à Rémy TESTON pour sa visite, à l’experte eSante Evelyne KRATZ, David KALFON, Pascal BECACHE et bien sûr l’éblouissante Marguerite BRAC DE LA PERRIERE ! Alors : #Bravo Jérôme LELEU, Julie BONNET, Anne Durand pour cette réussite ! #Merci à tous les intervenants et tous les visiteurs de notre stand que je ne cite pas ! #Merci aux personnes qui ont utilisé notre sac (poke : AXELLE MENU-BRANTHOMME 🙂) À très vite ! #eSante25
-
-
-
-
-
+8
-