Dec 2022 – Dec 2023 · 1 yr · New Delhi, Delhi, India · On-site

• Developed and tested futures and options trading strategies and analyzed the performance on data of 1080 days using Pandas.
• Optimized code to reduce runtime by 40%. Used Matplotlib to visualize and plot financial data.
• Developed pan-India trader database using MongoDB.
• Automation of data mining from NSE & BSE websites to retrieve useful data with 0.85 second latency.
• Automated summary generation of Introducers, capital management, and upload data to NSE as Project Manager for developing back-office software.
• Compiled and analyzed databases with more than 50 million data entries using Big Data Analysis tools like Dask, Apache PySpark.
• Developed and maintained data-driven solutions for credit management, resulting in improved risk assessment and customer experience.
• Worked in Machine Learning implementation and supertrend indicator analysis on historical data.

Oct 2022 – Nov 2022 · 1 mo · Remote

• Development of an automated surveillance system that utilizes live video streams from raspberry pi and processes them using AWS Kineses Video Streams and AWS Sagemaker.
• Utilized S3 buckets for storage of surveillance data.
• AWS SNS was used to notify the user of any unwanted personnel.

Feb 2022 – May 2022 · 3 mos · Jaipur, Rajasthan, India

• Optimized ML models and developed machine learning pathways for custom datasets and models (Automated Number Plate Recognition system).
• Model compression (using quantization and pruning) which reduced inference time from 250ms to 120ms; accuracy 92% in ANPR and 90.4% in YOLOv5.
• Worked on translating ngrok source code from Golang to Python to transfer and receive docker payloads via tunneling.
• Deployment of docker images for detection on edge devices (Jetson Nano & Raspberry Pi) with low compute power via SSH tunnel.
• Streamlined training for a number of datasets.
• Deployment of Nvidia and custom docker images on edge devices.
• Engineered a robust pedestrian detection system leveraging TensorFlow, PyTorch and computer vision, enhancing surveillance capabilities on edge devices, drones.
• Model compression and optimizations using quantization, pruning, Onnx, TensorRT

Sep 2021 – Jan 2022 · 4 mos · Remote

• Developed a dataset using web scraping and an intent classification model with accuracy 95% for a news recommendation system.
• Text analysis using Information Retrieval techniques like BM25 and data pipelining using Redis.
• Deployment of Question answering and Question generation using BERT, ROBERTA from Hugging Face for an interactive user experience.
• Fine-tuning Large Language Models (LLMs) for a news recommendation website; accuracy improved from 85% to 94%.

Jun 2021 – Aug 2021 · 2 mos · Work from home

• Successfully deployed a Flask-based website with an integrated machine learning model in the backend to automatically predict and display related graphs.

Apr 2021 – May 2021 · 1 mo · India

• Learned and implemented various libraries and modules of sklearn, Scikit, Numpy, Pandas, Tensorflow, and PySpark.
• Developed machine learning models for classification, regression, clustering, NLP, Computer Vision, and time-series analysis.

Feb 2021 – Mar 2021 · 1 mo · India

• Speech-to-text converter with GUI using Python: The project utilizes the SpeechRecognition module of Python programming language to convert sound or words spoken by the user into text and save it as a text file.
• RPC implementation using Python: Researched RPC (Remote Procedure Call) and its implementation using Python programming language. Used XML-RPC and gRPC to design and communicate between a server and a client.
• Graph plotter in Python A Tkinter UI-based Python project to display graphs of the marks of a student.