Photos and images have a lot of data encapsulated in them. Extracting them with a trained human eye can take considerable time and effort. Instead, you could rely on an alternative that does the work in half the effort applied by a human. We are talking about machine learning and artificial intelligence. Integrating both the fields to get necessary information off documents and images within seconds, and one company aims to do that in the most fashionable manner possible.
Docsumo is an artificial document and image analyzer that extracts information from documents, images, pictures and structures them into usable data. These could be invoices, KYC details, ID cards, and many more. Adding to this, the company helps in fraud detection whether an ID card is photoshopped or there are changes in its fonts. It’s quite a unique tool that saves hours of effort applied by humans in entering the information manually where it can be done automatically and in seconds. Speaking with Rushabh Seth, the CEO and Co-founder of Docsumo, Venture Mirror had a brief overview of its inception and other aspects of the company. Learn more through the interview illustrated below.
What is the startup / product / venture about? Give us a brief description of it.
Docsumo is a Document AI software that helps enterprises capture and analyze documents. It converts unstructured documents such as invoices, bank statements, ID cards, annual reports into structured data. Beyond data capture, Docsumo provides analytics that helps in automated decision making. It also checks documents for fraud such as photoshop layers and font changes. Enterprises can automate 90% of the workflow while saving up to 70% of operations costs with Docsumo.
How many co-founders are there? Please introduce them and their backgrounds.
Rushabh Sheth, CEO – Rushabh was the ex-Managing Director for Rocket Internet in Asia where he headed sales for an e-commerce venture. Post Rocket Internet, Bikram and Rushabh started a data science consulting firm where they helped multiple clients automate their processes. He has a Bachelor & Masters from IIT Bombay and MBA from IIM Bangalore.
Bikram Dahal, CTO – Bikram was leading business intelligence and data science for Rocket Internet in Asia where he met Rushabh. Bikram has built ranking and analytics solutions while at Rocket. Bikram has a Bachelor’s degree in Industrial engineering from one of the top colleges in Nepal.
How did you come up with the idea? What motivated you to do this?
Bikram and I have been working together for a couple of years before coming up with the idea for Docsumo. In our previous startup, we did accounting manually for a long time. We wasted days entering data on Quickbooks & Xero and reconciling bank statements with invoices. We tried different solutions in the market but nothing worked.
So we set out to solve the problem ourselves. Docsumo was born out of that desire to free founders, COOs, CTOs from manual data entry. We started with invoices as we knew from our experience that there is a need for such a product. But now, we help enterprises to process lending applications, bank statements, ID cards, tax returns, bill of lading, contracts, and many other document types.
Who is your target market? Why do you think your product will appeal to them?
Docsumo targets lenders and insurers who are looking to go digital. Docsumo already works with a few financial services in the USA and marque Indian customers such as BYJUs, PayU, Hitachi among others in India. The main value that customers get is that they are able to onboard customers in real time which would otherwise have taken them 2 to 3 days with manual processes.
Who are your competitors? How are you different from them?
There’s a lot of awareness and plenty of investment in AI, Machine Learning, Data Science, and related technologies. As a result, this domain is expanding with each passing year. We directly compete with Google Document AI and Amazon Textract and a few other niche players such as. Rossum.ai, Abbyy, and Kofax. Our customers like us over our competitors, because:-
1. Use case-specific pre-trained APIs -> Helps customers to quickly test and see success.
2. Analytics -> Helps customers to not just get structured data but also get attributes/data cuts which are required for decision making.
3. One-stop shop solution -> The ability to train other document types helps our customers. Customers frequently have said that they want one solution and not manage 3-4 different vendors for different use cases. This is why we are always working on making our New Doc Type faster to train, more accurate and more customizable.
4. Support -> I cannot emphasize enough how deeply we care about customers. Our customers have told us that they picked us because we were always available to solve queries while testing and help the customer solve their problem end to end. This is a clear differentiator for a complicated API product like ours.
What are the future plans with the product/startup? Any new features you are planning on?
Going forward, the company is building a suite of API products that helps lenders and insurers automate their operations end to end. We recently launched our free tool, Extract Tables by Docsumo, on Product Hunt.