We help you accelerate your digitalization journey
AgileDD is actively involved in research and development to build solutions driven by machine learning and deep learning. In particular, AgileDD is collaborating with several E&P companies to develop reusable customer solutions in the general areas of document information extraction, computer vision and time series prediction.
We welcome queries related to R&D and consulting in the general area of data science.
We have collaborated with a major E&P client to develop a method to extract lithological intervals, hydrocarbon shows and geological descriptions from composite logs and map these features to depth.
We have internally develop a deep learning based method to improve Optical Character Recognition for difficult documents with multiple text orientations.
In collaboration with several O&G operators and major contractors, we are developing a method to extract tabular data from unstructured document with machine learning combined with graph algorithms
We have internally developed Auto-Predict, an AutoML time-series library that non-data science experts can use to make accurate time-series prediction models
Publications and Presentations
Classification and metadata extraction using machine learning, NVIDIA Global Technology Conference, GTC 2019, Poster 9250,
Method and system for extracting, verifying and cataloging technical information from unstructured documents, Allowed (Approved) US Patent, Filed December 2016, Published December 2019
Extracting information from well logs using computer vision, Parminder Kaur and Jacques Micaelli, Professional Petroleum Data Expo, April 9 and 10, 2019, Westin Houston Memorial City, Houston, Texas
Classification and metadata extraction using machine learning, Parminder Kaur, Professional Petroleum Data Expo, April 9 and 10, 2019, Westin Houston Memorial City, Houston, Texas