Designing Voyance Vision on Voyance Cloud

Every day, businesses, both big and small, have to analyse tons of documents supplied by their customers during the compliance processes to ensure that these documents are legal, correct and have not been tampered with.

This can be a tedious task and takes a lot of time. The question we had to answer was:

How can we simplify this process, so that businesses can focus on more important tasks like improving product features and delivering excellent customer care?

The answer was Voyance Vision.

A computer vision solution that provides an automated, fast, accurate platform that is able to handle processing limitations such as low image quality or poorly scanned documents.

Voyance Vision is a part of Voyance Cloud that allows businesses to create and train models or use existing models created by our Data Engineers to extract texts from documents such as invoices, receipts, passports, drivers licenses, or other forms of documents.

Voyance Vision uses OCR technology to make this possible. OCR stands for Optical Character Recognition. It converts texts or information extracted to machine-readable data. These data can be edited or stored for further processing. Voyance Vision uses Artificial Intelligence and Machine Learning to make this possible.

Although OCR has a vast use case, the use case for this application would be to verify and approve legal documents and automate data extractions from these documents and store them for future use. This step is crucial in compliance processes.

About the Project

Voyance Vision is a cloud-based solution that helps businesses solve computer vision problems related to document and image analysis.

Voyance Vision v1 allows users to extract texts from images. Future updates will include PDF documents.

Product Manager: Jennifer Okafor

Product Designer: Sebiomo Anuoluwapo

Research Engineer: Odunayo Ogundepo

Frontend Engineers: Oluwanifemi Adeyemi, Isaac Bello, Victor Akinyemi

Backend Engineers: Temi Babs, Patrick Ojunde, Vicky Foukou

A Practical Problem Vision Solves.

The FinTech space is a fast-growing niche in Nigeria with each company having thousands of users. According to TechCabal, Kuda bank has about 650,000 customers, Paystack has about 60,000 customers according to TechCrunch, CowryWise has over 220,000 customers (TechCrunch), these are only a fraction of the companies in Nigeria.

As a part of the compliance process which ensures the authenticity of each customer and the overall integrity of the platform, customers need to submit any government-issued ID such as Driver’s license, Nation ID Card (NIN), Passport, Voter’s Card and or CAC Documents for businesses.

It can be a daunting task to manually go through the tons of documents submitted each day. These manual processes of extracting the needed data from these documents can be expensive, time-consuming, and error-prone.

The Vision Solution.

Voyance Vision is a solution that helps reduce cost and the time spent processing these documents without compromising on the accuracy of data extraction.

Voyance Vision gives companies the ability to train their models or use our pre-built models that have been trained by our Data Engineers.

Our Process

We have simplified the process from start to finish to enable our users to get started right off the bat.

First-time users are given a free 14-day trial period and a specific amount of API calls, which allows them to use our services without any commitments.

After the trial period, users are required to make payments to continue using Voyance Vision.

Users can easily get started using our prebuilt models which do not require training as our Data Engineers have taken time to train models that can be used immediately.

Files can be added locally, with the plan for future versions to support Google Drive and Amazon S3 Bucket.

Once the data have been extracted, they can be stored for future use in third party databases that we support. These are MySQL, MariaDB, Amazon Redshift and Postgres.

The Technologies We Used.

Our Engineers used:

• Tensorflow, an open-source library for machine learning, mostly used for deep learning, was created by the Google Brain team.
• Pytorch, also an open-source library for machine learning, is commonly used for Computer Vision and Natural Language Processing.
• OpenCV and
• Kubernetes, a software orchestration tool.

Challenges we Faced

The major challenge was identifying the specific solution we were going to offer at launch.

Because computer vision is so robust, you can do a thousand things with infrastructure like Vision. So my work was to cut out many parts to create what our “right-now” clients need.

Also, what we are building, an AI infrastructure is more complex than your everyday tech product.

Personally, I had to learn fast and have a clear picture of the product, to help the team with understanding it too. I am still doing the learning aspect because people are always coming to me with questions, and I should have the answers.

- Jennifer Okafor, Product Manager.

The disparity in the understanding of the problem was also a challenge.

This was tackled by sharing concise product documentation across the team, and having regular sessions with the product team, to avoid assumptions.

Formalization of the engineering tasks, going from product description to a precise engineering breakdown. Architectural decisions had to be made, options had to be weighed carefully to optimize the code for efficiency and scalability.

- Temi Babalola, Senior Software Engineer.

Working on this project, for me, was a learning experience.

Before this moment, I did not have any background knowledge in OCR technology, Computer vision, Machine learning and Artificial Intelligence.

I had to study and do research to develop an intuitive design that solves the problem. Working with the team on this project was amazing.

- Sebiomo Aanuoluwapo, Product Designer.

The Results We Got.

With Voyance Vision, our customers started securely extracting data from thousands of IDs, easily transferring the needed data to the right departments.

For one of our clients, a task that would have taken them 8 months was completed in days. All at a significantly lower cost!

With the infrastructure Voyance provided, this company got:

An automated document verification process.

A connection hub to extract, transform and load data from different sources for easy data management.

Intelligent business intelligence and behavioural analysis.

- Folasade Daini, Content Marketing.

Voyance Vision takes out the slow, manual, and error-prone system and replaces it with AI-powered infrastructure that cuts out expensive intermediaries and manual work. It also provides access to a connection hub for easy data management and a machine learning-based prediction tool.

All the tools you need are available to you in a single platform, helping you save thousands of dollars you’d have spent on multiple vendors.

Let nothing stop you from getting these results for yourself. Start saving your business time and money now. Begin your free 14-day trial now.