Iris.ai is searching for research partners!

About the company

Iris.ai is a Norway based start-up behind the award-winning AI engine for scientific text understanding. Using the latest Natural Language Processing (NLP) technologies, which systemize human text understanding for machines, we are building an AI Researcher to assist research processes such as information retrieval, hypothesis generation, relation extraction and cross-domain knowledge discovery. Our current tool, the Researcher Workspace, is releasing 75% of the work hours of researchers from tedious manual work, and our R&D team is building scientific knowledge graphs to help even more.

 

About the product

The Researcher Workspace is a platform for processing research documents, including a range of smart tools including search, filters based on corpus analysis and context, summarization and extraction of experiment data. The tool suite can be specialized on each client’s research domain without human training. It is being used in biotech R&D, post market surveillance in pharma and medtech, and material science competitor IP analysis - to mention some. All successful research results eventually become productized into this tool suite.

 

About the research project(s)

Our current and near-future research goals include (but is not limited to) (1) improving domain adaptation for word embeddings - which transforms and combines embeddings spaces through various techniques to understand their advantages and disadvantages; (2) knowledge graphs construction - which imputes word-embeddings information to enrich entity relationships on a knowledge graph; (3) entity extractions from tabular and graphic data in scientific literature - which enriches the system with non-textual data for better representation of research results; (4) abstractive summarization and its adaptation - which creates better representations of document clusters in various domains; (5) embeddings and summarization evaluation suite - which allows a systematic overview of performances for our core algorithms.

 

Current collaboration partners

In a soon-ending BIA/IPN project, we collaborate closely with UCL, RISE and Core.ac.uk/Open University. We also have past and current collaborations with other academic institutions such as Chalmers, Oxford and Stockholm University.

 

In search for new collaborators

We are currently looking for:

  • Academic groups within NLP or adjacent fields interested in joining our in-progress IPN application prior to September 14th as an FoU partner.
  • Corporate partners interested in joining the same IPN application as a tester/evaluator of smarter tools for their R&D. You should be operating in MedTech/Pharma or Material science/Chemistry, with a need to better manage knowledge from large collections of research documents.
  • And of course any other opportunities for short or long term collaborations.

Get in touch!

We’re currently deep in the writing of the IPN application, so get in touch right away to join us, on founders@iris.ai

Publisert 19. aug. 2022 15:03 - Sist endret 19. aug. 2022 15:03