Srikanth here, co-founder of DocuSutra (https://docusutra.com). Excited to show our product, which helps with customizable data capture for documents like invoices, bills, purchase orders etc.
As part of our previous work assignments, we were helping our clients with data extraction from documents like invoices, lading bills etc. We relied on cloud based solutions like GCP Vision, AWS Textract etc to do the heavy lifting for us. One of the key drawbacks we noticed is the lack of ability to accommodate the custom requirements like specifying additional custom fields, table columns etc into their models. This resulted in writing quite a bit of glue code to address the custom requirements. And additionally, whenever there are format changes, we ended up reworking the glue code to accommodate them.
With our current solution, we are enabling the end user to teach the system to act as per the requirements. All that needs to be done is to tag the relevant portions of the documents with appropriate labels. And the system is sensitive enough to learn with a very few (~5) document samples. In case of any format changes, all that needs to be done is to teach the system with the new documents. No further need of developer intervention.
We tried to make DocuSutra as easy to use as possible to get started with. Currently, it is in beta and free to use. In any case, there will always be a free tier (no credit card requirements).
Srikanth here, co-founder of DocuSutra (https://docusutra.com). Excited to show our product, which helps with customizable data capture for documents like invoices, bills, purchase orders etc.
As part of our previous work assignments, we were helping our clients with data extraction from documents like invoices, lading bills etc. We relied on cloud based solutions like GCP Vision, AWS Textract etc to do the heavy lifting for us. One of the key drawbacks we noticed is the lack of ability to accommodate the custom requirements like specifying additional custom fields, table columns etc into their models. This resulted in writing quite a bit of glue code to address the custom requirements. And additionally, whenever there are format changes, we ended up reworking the glue code to accommodate them.
With our current solution, we are enabling the end user to teach the system to act as per the requirements. All that needs to be done is to tag the relevant portions of the documents with appropriate labels. And the system is sensitive enough to learn with a very few (~5) document samples. In case of any format changes, all that needs to be done is to teach the system with the new documents. No further need of developer intervention.
We tried to make DocuSutra as easy to use as possible to get started with. Currently, it is in beta and free to use. In any case, there will always be a free tier (no credit card requirements).
Thanks and looking forward to your feedback!