What does it take to be a Digital Insurer? Well one of the basics is to have a deep and a sound understanding of Machine Learning (ML) and Artificial Intelligence (AI) with the capabilities to actually develop these for ‘insurance’ case studies.
Our main AI product is a set of several AI modules, we design very robust AI modules for individual requirement. While at the same time, users/products can do multiple analysis efficiently.
MIC used Auto-Machine Learning (Auto-ML) approach in our AI product, where Auto-ML decides the best ML Algorithm. We used best ML Algorithm as suggested by Auto-ML which enhances our product capability.
We see other companies develop generic AI products, then focus on insurance domain. However, our products are designed specially for Insurance case studies.
Typically companies say that they use AWS or Google or TensorFlow. We don’t use single ‘AI’ from the likes of AWS or Google.
What we do is use a ranges of tools and our own algorithms and code, the core being made up as follows:
TensorFlow is an end-to-end open source platform for machine learning. It has a comprehensive, flexible ecosystem of tools, libraries and community resources that lets researchers push the state-of-the-art in ML and developers easily build and deploy ML powered applications
Keras is a high-level neural networks API that works as a layer.
Deep Learning (also known as deep structured learning or hierarchical learning) is part of a broader family of machine learning methods based on artificial neural networks. Learning can be supervised, semi-supervised or unsupervised.
Natural Language processing (NLP) is a subfield of computer science, information engineering, and artificial intelligence concerned with the interactions between computers and human (natural) languages, in particular how to program computers to process and analyze large amounts of natural language data.
Optical Character Recognition or optical character reader (OCR) is the mechanical or electronic conversion of images of typed, handwritten or printed text into machine-encoded text, whether from a scanned document, a photo of a document, a scene-photo (for example the text on signs and billboards in a landscape photo) or from subtitle text superimposed on an image
Machine Learning (ML) is the scientific study of algorithms and statistical models that computer systems use in order to perform a specific task effectively without using explicit instructions, relying on patterns and inference instead.
Tensor Flow and Keras takes maximum memory and computing time, it is very big challenges for companies to deploy deep learning-based application. We over come this using our embedded application, this takes less memory and less computing time (0.5-1.2 sec per image) and can run on any operating system.
Our AI product allows us to focus on accuracy, for our use cases and successfully identifies body parts, damage, colour with user input images with 80-90 percent accuracy. Furthermore, we developed robust algorithm to calculate damage severity with very good accuracy 75-80 percent. This is growing as more data is applied and as we add features.
Our AI product takes less computing time with highly mechanized deep learning neural network architecture, this enables us to provide faster solutions suitable for insurance. Additionally we deal with error tolerance and to handle outlier/noise. This helps support multi-tasking with error less environments.
We use our team and they have their own definitions of our licensed tech.