DataSeer, Inc.

A SaaS application that detects, extracts and digitizes data from industrial diagram, data sheet and form PDFs

We currently have a handful of enterprise EPC customers in the oil & gas sector, i.e., Worley, Technip, Suez, Stork, and many smaller customers / accounts. We are in Year 2 of being out to market, have renewed all our customers from Year 1 and are in the process of branching out beyond EPC and oil & gas into other sectors like construction, water & recycling, HVAC and others.

Our goal is to accelerate information extraction from non-digitized technical engineering diagrams that is currently done manually. The manual process is costly in terms of time and money, with potential impact on performance, quality, and safety. Our product allows for time savings in the extraction process while keeping the engineer in control. Our cloud-based SaaS solution leverages Computer Vision algorithms and Machine Learning models and allows industrial engineers to work more efficiently, transforming engineering work to a semi-automated and technology-accelerated effort.

Our web-based SaaS solution uses computer vision algorithms and machine learning models to learn which symbols are relevant, extract any related text attributes and connect symbols via associations. We offer a deep net model that automatically detects most common symbols in process engineering diagrams. We also allow users to define their own custom symbols to detect via a human-in-the-loop machine learning paradigm. Our tool has an intuitive verification workflow that makes data review easy and fast. The underlying technology gives the user flexibility to process different kinds of diagrams.

We have a SaaS subscription revenue model with a minimum 12-month subscription. We have several pricing model options, depending on customer needs, i.e., seat-based for SMBs vs consumption-based pricing for enterprise. We offer paid pilots to allow customers to evaluate the tool and establish proof-of-value before they subscribe.