![](http://www.poornimasuresh.com/wp-content/uploads/2020/01/Streamlit.png)
![](http://www.poornimasuresh.com/wp-content/uploads/2020/01/Screen-Shot-2020-01-13-at-6.16.54-PM-1024x586.png)
Python Web App
This app was written using the streamlit library, and features a Keras/TensorFlow machine learning demo that predicts the chance of becoming diabetic based on various lifestyle parameters. The app is deployed on a Docker container on AWS Fargate.
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-1a.jpg)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-1b.jpg)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-1c.jpg)
DermaDetectâ„¢
This mobile application leverages AI to detect skin cancer. The app displays a live video stream along with the probability distribution of benign vs. cancerous. By applying Applied transfer learning on Inception V3 model, I was able to boost the accuracy of the prediction.
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-2a.jpg)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-2b.jpg)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-2c.jpg)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-2d.jpg)
Autonomous Vehicular Traffic Management
Our team designed an algorithm for a future where the majority of vehicles are networked and autonomous. By simulating a “no-stop intersection” where vehicles space themselves apart such that two opposite streams of traffic can pass through an intersection simultaneously without stopping, we saved travel time by 11%.
Languages
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/home-2-1024x666.jpg)
- Python
- Unix shell (bash)
- C++
- Swift (iOS)
- Java (Android)
![](http://www.poornimasuresh.com/wp-content/uploads/2019/08/portfolio-2d.jpg)
Contact Details
Replace this text with descriptive copy to go along with the card image. Then add more blocks to this card, such as buttons, lists or images.