Himanshu Pareek

Dynamic and results-oriented Machine Learning Engineer with 1.5 years of experience in developing, deploying, and optimizing advanced machine learning models. Demonstrated expertise in deep learning, natural language processing (NLP), and image processing using tools like TensorFlow, Keras, and Python. Proven ability to collaborate effectively with cross-functional teams, contributing to successful research initiatives and business outcomes in high-impact projects with organizations such as ISRO and A3logics.

Associate Software Engineer

A3logics Pvt Ltd | Oct 2023 – Present

• Developed an EDI translation system using LSTM-based Encoder-Decoder architecture, achieving 98% accuracy.
• Managed data preparation, including tokenization, embedding, and NER, to ensure high-quality input for optimal model performance.
• Designed a Retrieval-Augmented Generation (RAG) system with LangChain and LLMs/SLMs, automating data retrieval to enhance efficiency.
• Coordinated with front-end and API teams for seamless model deployment on the server.

Malaviya National Institute of Technology (MNIT) | May 2023 – Sep 2023

• Created a CNN-based U-Net Semantic Segmentation model, achieving an F1 score of 0.91 on satellite images of Jaipur.
• Conducted data labeling and preparation using LabelBox, enhancing dataset quality for model training.
• Applied noise filtering and clustering techniques on LiDAR data for building height retrieval, achieving an 11 cm accuracy.
• Presented research findings at ADCIS 2023, contributing to advancements in remote sensing.

Dual Degree B.Tech + M.Tech in Information Technology

University of Rajasthan, Jaipur | 2018-2023

• GPA: 5.63/6