I have dived deep into technologies such as Front-end and Back-end Development, and IoT with Arduino and Raspberry Pi 2, however, I know that my interest lies in AI/ML.
In 12th grade I wrote a research paper - "Managing Congregations of People by Predicting Likelihood of a Person being Infected by a Contagious Disease like the COVID Virus" that got published in IEEE CCEM 2020 and won the Best Short Research Paper after I presented it in the conference.
I created an interactive dashboard for a startup called AarogyaAI using Streamlit which I integrated into the AWS S3 bucket to display and save results and reports of TB patients.
I have completed projects such as Face Recognition and Image Inpainting as a first-year undergraduate at SRM Institute of Science and Technology.
I have mentored students in McCarthy Lab, Next Tech Lab in the field of AI/ML.
Projects
Public anomaly detector
Checks the security footage for any public security anomaly taking placeJavaScript, Flask, TensorFlow, Keras, OpenCV, Python, TypeScript, Convolutional neural network (CNN), Long short-term memory (LSTM), nextjsSkills
Experience
- AarogyaAI - Data Science InternJuly 2020 - September 2021
Weeding out anomalies in the data of TB mutations and verifying them against another dataset. Skills learned - Pandas and working with dataframes.
To automate the calculations of Sensitivity, Specificity, Positive Predictive Value, and Negative Predictive Value by using the results from two ML models stored in CSV files.
Created a dashboard/frontend to display datasets and results stored in AWS S3 bucket. Skills learned - Boto-3, Streamlit.
- Upthrust - AI Prompt Engineering InternJune 2023 - July 2023
Integrated PageSpeed Insights API and Screaming Frog API to produce website issues and used LLMs like GPT-4 in
LangChain to fix those issues.
Scraped the PageSpeed Insights documentation and created a vector store containing chunks of embeddings which was used to
query those issues and find the fix.
Linked the Meta Marketing API with GPT to create a chatbot which provides recommendations to fix aspects of ads where it’s
going wrong.