The adoption of AI/ML in healthcare has radically changed risk profiling, diagnosis, and underwriting in medical insurance. To understand how Fedo is using AI and machine learning, we got in touch with Arun Mallavarapu, co-founder & CTO of Fedo.
Founded by Arun Mallavarapu and Prasanth Madavana in 2017, Fedo is a health-tech startup that uses an AI-enabled system to generate a health score which is more like a Trans Union credit score for finance. It has built an AI/ML platform that quantifies an individual’s risk for various diseases and his/her propensity to claim over the next few years.
Fedo Score is a holistic indicator of the future health risks of an individual. It also offers insights on the likelihood of incurring a medical expense in the next few years by taking the individual’s demographics, lifestyle, and habits into account.
The Fedo health score was built by medical professionals and data scientists using 250+ medical studies, 2000+ plus quality controlled and academic and research documents from all over the world and analysing over 50 million global health records and 1.5 million claims data. The risk calculated through Fedo’s algorithms is used to determine underwriting decisions in real-time.
Facial Recognition system takes a facial image as an input and predicts features like BMI (body mass index), smoker or not, age and gender.
What’s the differentiator
Mallavarapu said, “Predicting diseases with the help of AI & ML has been hugely researched upon. Many health analytics companies predict diseases and provide a solution to patients. FEDO predicts the risk of six chronic diseases and quantifies the risk into a health score through different AI/ML algorithms which sets it apart from all the other products in the market. It also segments the FEDO score into three different categories – unhealthy (<400), moderately healthy (400-600), healthy (>600) so individual health can be better mapped. In the age of big data, where data is readily available all around, FEDO has got the skilled resources, capacity, and channels to leverage the data and process it to invent innovative products.”
Use of AI & ML at Fedo
The company uses a wide variety of AI/Ml techniques ranging from classification and regression models to convolutional neural networks. “We have a stronghold on the AI/ML side and are aware of when to use a specific technique. For instance, we use transfer learning to learn the lower-level features in computer vision use cases so that the data required to build efficient models reduces drastically. In addition, we use a variety of machine learning algorithms for classification/regression problems. Our uniqueness lies in the fact that we don’t just pull an out-of-the-box model and execute it on data, rather we go under the hood of the specific model we are implementing and tune the mathematical and statistical components as they relate to our problem,” said Mallavarapu.
The core models of Fedo are built using MATLAB and Python. Also, the models are typically consumed through APIs hosted on AWS and Azure cloud servers. The company uses docker extensively and follows standard dev ops and agile processes for deployment and development, respectively.
Mallavarapu said Fedo will help in quantifying the health risk of people and will make the underwriting process much easier and faster. The five year plan includes-
- Generating awareness among the people about the importance of health & life insurance.
- Awareness about benefits of new technology like AI/ML in healthcare.
- Making Fedo Score the gold standard of health score globally.
- Collaborating with a greater number of insurance companies for retail and group underwriting process across nations.
- To emerge as a leader in the health-tech industry.
Join Our Discord Server. Be part of an engaging online community. Join Here.
Subscribe to our NewsletterGet the latest updates and relevant offers by sharing your email.
A Technical Journalist who loves writing about Machine Learning and Artificial Intelligence. A lover of music, writing and learning something out of the box.