I love to write poems, tell stories and play games. Poems about people. Stories about graphs. Games about the unreal.
Since you are already on my portfolio, you will get to know everything regarding my technical skills and interests. So I'll tell you something you won't find on my resume. Well then, I'll tell you how I initially made friends in college. You see, people love talking about themselves and get excited when someone says that they would like to write something for them. And that's exactly what I did. I strike up a conversation and then let them know that I'd love to write a poem for them if they could tell me somethingthemselves. I take what they tell me, write a poem using the information and we become friends. I still think it should not have worked. But, it works so it works, I guess? And it helps a lot when you are trying to understand the other person.
Data for each month of the year 2019 for a sales company was generated and statistical and association analysis was performed. Multiple sales questions were answered. Interactive graphs were plotted using Plotly.
Analysis of the Wisc-bc-data from kaggle on a jupyter notebook. Feature engineering and statistical analysis was performed. Machine learning algorithms were employed to detect the type of cancer.
Detection of the severity of heart failure in patients in a scale of 0-4 was performed using artificial neural networks in Keras. The model was trained for 100 epochs with 84% accuracy.
A dataset with information about people was taken and analysed to predict whether the person would earn greater than or less than 50,000 USD (Hopefully a month) or not. The inconsistent data needed a lot of preprocessing.
A Real time threat detection system build for Kerela Police during the hackathon HAC'KP. Face recognition was performed using opencv and Nvidia Jetson Nano for camera integration to detect faces from the police database.
Future sales were predicted using a regression model for a custom dataset. The rating, sales in the first and the second month were used as input and the saled for the third month were predicted. A flask app was build by deploying the pickle model into heroku.
A Python command line app (CLI) that installs all required modules to any python environment with the help of the OS and Sys module. The user's desired domain, locations and environments can be specified at execution time using the argparse library.
Systems Engineer in the Data and Analytics (DNA) unit of Infosys, Hyderabad. I oversaw ETL operations of over 2000 workflows and databases using Informatica. Administered mappings, sessions and workflows for busniss applications. I was responsible to fix data related issues in the applications.
As a data analyst intern I was first trained in the basics of python, mathematics and statistics and finally the anaconda distribution system. I also worked on multiple real-world datasets and built machine learning models to solve the problem statement after exploring the data.
As a core team member of codezoned, I've had the opportunity to lead the Web dev team to build the official Codezoned website. I've also contributed to multiple other projects in codezoned like the Github Auto-inviter and led the tech team in the HAC2020 hackathon with MLH and Github.