Work
Experience
-
PTC Software
Software Analyst - AI Engineer
Pune, India - Hybrid- Developed an AI-assisted product support solution using Generative AI and Retrieval-Augmented Generation (RAG) techniques, enhancing internal processes and productivity while dynamically automating data updates. Now in production, this solution has significantly reduced the workload of two teams through the use of Generative AI.
- Delivered a Data Engineering Proof of Concept (PoC) that transformed real-time data into two augmented layers using delta tables and Databricks, enabling microservices to effectively extract data and successfully addressing challenges with large datasets; this PoC has been transitioned to production use.
- Developed an ONNX model scoring feature using Java and Python with the ONNX Runtime (ORT), which extended the platform's AI capabilities and facilitated the seamless integration of diverse machine learning models into the existing architecture.
- Contributed to the ThingWorx Analytics R&D team by enhancing the platform's ability to handle large-scale enterprise data, which involved updating the analytics library, improving overall performance and scalability, and addressing bugs in the existing codebase.
-
Quantiphi
Machine Learning Engineer - Intern
Mumbai, India - Remote- Developed a License Plate Detection and Recognition Capstone project using YOLOv4, CNNs, and OpenCV, gaining practical experience in machine learning and computer vision while delivering an effective solution.
- Contributed to the development of a Visual Search Engine using Vision Transformers and implemented a vector database search with FAISS to optimize performance, enhancing knowledge of cutting-edge ML technologies and deepening proficiency in CNNs, OpenCV, and modern ML frameworks.
-
Magna International
Software Development Engineer - Intern
Pune, India - Onsite- Built a comprehensive data analysis and prediction tool using PyQt5 for a major client, enabling efficient data visualization, predictive insights, and anomaly detection to improve decision-making processes.
- After successful tool development, fulfilled a major requirement by adding map data integration to visualize location-based data, further enhancing the tool's geographic analysis and prediction capabilities.
-
IOTians Global Innovations
Frontend Developer - Intern
Pune, India - Remote- Creating client-centric websites/webpages, tailored to their specific needs, employing either Forward engineering or Reverse engineering methodologies, based on contextual requirements.
Personal/Research
Projects
Hover on the cards and click to explore the project demo or code.
Explore my work, including hackathon projects, achievements, publications, and personal projects tie right into these, showcasing creativity and skills I have honed along the way.
DocuQuery-AI 📑 ⏯️
This project is a Gradio-based web app that allows users to query PDF documents using a Retrieval-Augmented Generation (RAG) approach, leveraging large language models to extract information. Users can upload PDF files, create a vector database from them, configure the settings for a chosen LLM, and interact with a chatbot to retrieve information from the documents in response to user queries.
Weather Prediction Project 🌦️
This project utilizes machine learning and deep learning to predict weather conditions in a country based on historical data. It includes data preprocessing, feature engineering, GRU model training, and evaluation.
Newsify-AI 📰 ⏯️
This project is a Gradio-based web app that fetches and summarizes news articles using the NewsData.io API and AI-powered models from Hugging Face. It allows users to specify a query, country, language, category, and the number of pages to fetch, and then generates brief or detailed summaries of the articles retrieved.
Deepfake Detection 👨
This project is focused on detecting deepfake images using a deep learning model built with TensorFlow and Keras. The model is trained on a dataset containing both real and deepfake images.
PromptMate-AI 🫂 ⏯️
This project is a Gradio-based web app that generates detailed AI-powered responses to user-submitted questions using the Meta-Llama-3-8B-Instruct model from Hugging Face. Users can input a question and provide a custom prompt to guide the AI's response, and the app generates a detailed, contextually aware answer based on the provided inputs.
Ship Detection from aerial images 🚢
This project implements a YOLOv4-based ship detection system for aerial imagery. It addresses challenges like noise, scale variance, and data imbalance using advanced techniques such as CSPDarknet53, Spatial Pyramid Pooling, and Mosaic Augmentation.
LangTrans-GUI ✍
LangTrans-GUI is a Python-based graphical user interface (GUI) application designed to provide a seamless and user-friendly experience for translating text between multiple languages. Utilizing the powerful Google Translate API, LangTrans-GUI enables users to quickly translate text from one language to another with just a few clicks.Achievements, Awards
Publications
Hover on the cards and click to explore the certificates.
