Software Engineer with a Master’s in Data Science and strong foundations in Java, Python, full-stack development, and AI/ML. Experienced in building scalable backend systems, REST APIs, and deploying cloud-native applications with Docker and AWS. Skilled in developing intelligent, data-driven solutions using ML, RAG, and predictive modeling. Passionate about combining software engineering and machine learning to solve real-world problems and drive innovation in fast-paced tech environments.
Pune, Maharashtra, India
→
Summary
Contributed as a Full Stack Developer Intern, enhancing chatbot functionalities and optimizing backend operations for a leading automotive business services platform.
Highlights
Spearheaded critical feature enhancements for Jigyasabot's chatbot, integrating multilingual support, Excel report automation, bot template duplication, and RBAC-based controls, which significantly amplified platform scalability and user adoption.
Integrated a VADER-based sentiment analysis module to enable intelligent feedback interpretation and generate actionable insights, improving user experience and data-driven decision-making.
Engineered a robust shell script to serialize Airflow DAG execution for 1,100 daily files, effectively minimizing GCP compute overhead and eliminating resource contention through optimized parallel scheduling.
Awarded By
Graduate Aptitude Test in Engineering
Achieved a score of 446 on the highly competitive GATE examination, demonstrating strong foundational knowledge in engineering and computer science.
Java, Python, SQL, Shell Scripting, FastAPI, Django.
MySQL, MongoDB, FAISS, Docker, Git, Linux, AWS, GCP (Google Cloud Platform), Airflow.
Jupyter Notebook, Google Colab, Postman, VS Code, GitHub.
OOP, SOLID principles, RAG (Retrieval-Augmented Generation), Prompt Engineering, Agile Methodologies, Machine Learning, Deep Learning, Model Evaluation (F1, Precision/Recall, ROC-AUC), Transfer Learning, Natural Language Processing (NLP), Semantic Search, Vector Databases.
Communication, Teamwork, Initiative, Knowledge Sharing, Problem Solving, Strategic Thinking.
HTML, CSS, JavaScript, Flask.
CNN, Tensorflow, Scikit-learn, NumPy, Pandas, LLMs, Matplotlib, Sentence-BERT, Gemini LLM, Cohere LLM API, BERTScore, Word2Vec.
Data Extraction, Data Transformation, Data Visualization, Matplotlib.
→
Summary
Designed a FastAPI-based semantic search system for educational PDFs using Sentence-BERT and FAISS for efficient vector-based retrieval. Integrated Cohere LLM API to provide intelligent fallback responses and built a minimal frontend interface for smooth user interaction and contextual output display.
→
Summary
Implemented and evaluated image classification models including ANN, CNN, and SVM on the CIFAR-10 dataset. Leveraged TensorFlow and scikit-learn for model development, training, and performance evaluation, gaining hands-on experience with deep learning and traditional ML techniques for visual data analysis.
→
Summary
Developed an interactive browser-based 2048 game using HTML, CSS, and JavaScript, featuring dynamic tile movement, real-time scoring, and responsive UI for optimal user experience across devices. Integrated keyboard event handling to enhance interactivity and user control. Containerized the application using Docker and deployed it on AWS Elastic Beanstalk, enabling automated scalability and streamlined deployment workflows.
→
Summary
Developed a Retrieval-Augmented Generation (RAG) chatbot using Gemini LLM, Serper API, and India Kanoon API to provide accurate legal responses grounded in real-time web content and legal documents. Designed a "Deep Think" reasoning engine for modular, explainable multi-hop legal reasoning. Achieved 92% QA accuracy on an Indian legal benchmark using a BERTScore-based evaluation pipeline.