Sign In

Yiannis Kiachopoulos

CEO, Co-Founder at Causaly

Yiannis Kiachopoulos is a visionary in the field of biomedicine, with a focus on enhancing human cognitive abilities through advanced technology. As the co-founder and former CEO of Causaly, he has been pivotal in revolutionizing the way knowledge is accessed and utilized in the healthcare industry. With a diverse educational background that includes a Master of Business Administration in General Management from The Hong Kong University of Science and Technology, Future Studies from Singularity University, and Computer Science from FAU Erlangen-Nürnberg, Yiannis brings a unique blend of business acumen and technical expertise to his endeavors. Additionally, his fluency in Japanese showcases his commitment to continuous learning and cultural understanding. Prior to his role at Causaly, Yiannis served in key positions at reputable organizations such as Detecon International GmbH and Accenture GmbH, Germany, where he honed his skills in strategy, transformation, and information technology consulting.

Highlights

Aug 6 · linkedin

🔍 Why is GenAI on Scientific Information so hard? Scientists who are trying to understand disease biology, identify targets and innovative hypotheses normally are looking for uncommon knowledge, for the needle in the haystack. This means that the scientific questions asked are what can be seen as edge cases in a scientific search and answer engine (see image below). This “long-tail’ in Search is what makes developing reliable, transparent and robust GenAI Copilots for scientists hard.

While Enterprises can develop PoCs quickly, production-grade systems need to cover niche questions in the long-tail. Solving for this is a tedious, long process by: +Retrieval via controlled ontologies for reproducibility and transparency +Retrieval via semantic similarity for increasing sensitivity +Grounding results into a Knowledge Graph for transparency (Graph RAG) +Human-controlled calibration of Information retrieval +Re-Ranking of results to ensure top 10 results are the best match +Query reformulation to account for zero-results, and predictable search

  • …

The list goes on and on… Developing Information Retrieval and Ranking pipelines working on heterogenous scientific information is a multi-year process, one that we have invested in over the past 6 years. This, and our high-precision Knowledge Graph are the foundations for creating our robust scientific Copilot that works on external as well as internal scientific data. Companies should invest in robust foundations if user adoption (and therefore ROI) is a key outcome of introducing these new technologies in Life Science organizations.

Aug 15 · linkedin

🌎 🔬 Thrilled to unveil our Global AI R&D Conference on Oct 26/27 in #NewYork with great speakers from Takeda, Novartis, Meta, Roche and more to be announced soon. Join us to be part of the community discussing scientific innovation and the transformational shifts happening in biomedical research, AI and Drug Discovery and Development.

🔗 Register here: https://hubs.la/Q01-Xvg90

#AI #Research #DrugDiscovery #BiomedicalResearch #NewYork

Yiannis Kiachopoulos
Add to my network

Location

London Area, United Kingdom