Frequently Asked Questions


We have curated a few questions which may be on your mind. If you have any questions that are not in the list below, please write to us at and one of our experts will be in touch with you. is a bouquet of solutions. As the use-cases of table extraction, summarization and classification have many variants required by the clients, the base variant of the modules works for a specific set of inputs. The good news is we can customize it to your requirement. We have done successful implementations to our customers.

The solutions are not domain-specific, it can cut across all domains such as Banking, Retail, E-commerce, Manufacturing, Education, Healthcare and Lifesciences . What we care about are the formats of input, output and the training data for some modules. Otherwise, the solutions work for any domain if we get enough training data for summarization and classification.

It is based on a subscription model for the table extraction module. In some cases, we customise that model to give the precise output required by you.

For summarization and classification, a one-time service fee + product fee applies. An initial training functionality has to be configured. This typically ranges from 1 to 1.5 months where our data scientists tune the product to your data.

The product’s preferred mode of deployment is cloud. The application runs from the cloud and will fetch the data from databases. It can be customized for fetching data from on-prem databases and APIs. In such cases, an on-prem installation can be customized for the clients.

Zero maintenance cost, as long as the input and output format remains the same. We can make some changes in case of minor customizations, for which a service fee will be levied. A small maintenance cost is involved for server level adjustments.

The modules are off the shelf. They are readily usable if your inputs match our product’s input pipelines. However, 1 to 2 months of assistance from our team will be required if you are looking for a specific output.

The use-cases we have solved using TM are complex ones like Tables Extraction, Summarization and Classification. We have used a combination of ML, Ai, Text algorithms to solve these problems which usually cannot be solved. In addition, TM can be customized to your specific needs

Yes, tables from images can be extracted from images of different formats like jpeg, png, tiff and more. Our product also works in the cases where the pdfs and documents have embedded images. This is one of the distinguishable features. We have combined DL, OCR and popular table extraction methods to create a robust product.