Projects

Automation using AirOps 🔗 https://app.airops.com/public_apps/a5b053ed-3a94-4fc8-96cc-cec1878a62b7/run New updated workflow with much richer data : 🔗 https://app.airops.com/public_apps/cf4a3c94-36e1-4efa-a4dd-51a37862921e/run About Me : Ever since starting my master’s degree, I’ve been fascinated by large language models (LLMs) and how they’ve changed the way we learn. Tasks that used to take at least an hour, like getting the gist of a research paper, now take me about 10 minutes. I still explore the details, but LLMs make it much easier to understand the main ideas quickly....

Expand the dashboard for a detailed view data source : https://data.lacity.org/Public-Safety/Crime-Data-from-2020-to-Present/2nrs-mtv8/about_data What steps did I take to integrate the data source? ✔️Optimized Data Fetch with SoQL: I created a Dataflow to act as a middleman between the LAPD’s web API (SODA API) and my dashboard. But instead of grabbing everything, I used SoQL filters within the API call to snag only the columns I actually need for my analysis. This keeps the data transfer lean and minimizes the load on LAPD’s servers....

Abstract This study investigates the multifaceted nature of crime in Los Angeles between 2020 and 2023 by analyzing official LAPD data to uncover intricate patterns and nuances across temporal, spatial, demographic, and crime-type dimensions. Spatial analysis identifies high/low crime areas and hotspots through geographic mapping. Victim demographics are correlated with crime types, locations, and times. Prevalent crimes are categorized, and their patterns are assessed in relation to reporting times and proximity to police stations....

Linkit is a file sharing web application built on Flask and hosted on AWS. Seamlessly integrated with AWS services, offering User Authentication, file storage, dynamic prefix creation. check it out here

Previous Next     / [pdf] View the PDF file here.

Created a multifunctional conversation analysis application during a hackathon, offering audio input, playback, transcription, download capabilities, as well as summarization and sentiment analysis functionalities. check it out here

Developed and hosted a Streamlit Web Application on Streamlit cloud for Document Question Answering LLM using Cohere API. The user can upload txt, csv, pdf files. Ask questions on the document uploaded. Options to modify the temperature and tokens of the model.

Krishna Pranay Angara | Ramasai Badam | Nikhil Balachandra | Vishnu Koppera

Abstract Hand sign detection is an important concept for helping the challenged communicate. The project deals with the mentioned problem and uses deep connected neural networks for detection of hand signs and recognition of English alphabets. The wide spread of American Sign Language(ASL) was used for detection. Mediapipe framework was used for recognizing and extracting the keypoints through real time hand detection and saved in csv files which were fed to neural network for training....

Embarking on the journey of hosting a static website is an exciting endeavor, especially when you’re determined to harness the power and scalability of AWS. Throughout this guide, I’ll share my hands-on experience of successfully hosting a static site on AWS S3, shedding light on the challenges I encountered and the strategies I employed to overcome them. Acquiring a Domain from Google Domains The journey began with a simple yet crucial step: securing the perfect domain name through Google Domains....