Predicting Length-of-Stay & Risk of Mortality for ICU Patients

Pangaea Data Awarded Top Tier Co-Sell Partnership Status by Microsoft

Pangaea Named ‘Digital Solution of the Year’ by UK Government’s Department for International Trade

Global Pharmaceutical Proves Pangaea’s AI

Go back

Presenting “Novel AI to Automatically Generate Textual Narratives for Regulatory Reports” at Pistoia Alliance 2022

2:00 - 3:00PM ET (6:00 - 7:00PM GMT) November 2nd, Boston, MA, USA

Venue: Hilton Boston Back Bay, Boston, MA, USA

Presentation: PIES – Novel AI to automatically generate textual narratives for regulatory reports, based on data from experiments

This presentation will demonstrate Pangaea’s intelligence Extraction and Summarisation (PIES), a neural architecture, which uses a unique data augmentation procedure to address data sparsity (training with only 200 examples). Coupled with the copy mechanism, PIES ensures model interpretability and precision of values, which appear in the output (textual narratives). PIES is generalisable for various input datasets. The outputs are validated by human experts which allows efficient model improvement with minimal human effort.

Global pharmaceutical companies have demonstrated PIES novel methodology, which solves a challenging task of automatically generating textual narratives for regulatory grade reports in the medical and scientific domain for the first time. This has further shown that PIES is generalisable for various sets of input data to automatically generate assay validation, PTPK, Toxicology, clinical study reports and alike. Through PIES unique approach we have addressed the fundamental challenges of data sparsity and precision of values in text. PIES has also shown to drastically reduce the human effort normally required to improve the performance of such models through its unique Human-in-the-Loop protocol.

Schedule A Meeting


Dr. Michael Wu
Sr. Informatics Analyst
Genentech Inc.
Dr. Jingqing Zhang
Head of AI
Pangaea Data


Dr. Vibhor Gupta
Director & Founder
Pangaea Data