Harshit Samani Email & Phone Number

𝗔 𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗘𝗻𝘁𝗵𝘂𝘀𝗶𝗮𝘀𝘁, 𝗽𝗮𝘀𝘀𝗶𝗼𝗻𝗮𝘁𝗲 𝗮𝗯𝗼𝘂𝘁 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗮𝗻𝗱 𝗰𝗼𝗺𝗺𝗶𝘁𝘁𝗲𝗱 𝘁𝗼 𝘀𝗼𝗹𝘃𝗶𝗻𝗴 𝗿𝗲𝗮𝗹-𝘄𝗼𝗿𝗹𝗱 𝗽𝗿𝗼𝗯𝗹𝗲𝗺𝘀!- I am a skilled Data Scientist and Machine Learning Engineer with over 2.5 years of experience working on Computer Vision and NLP-based problems. I also worked on MLOPs and MLInfra-based challenges. I have exposure to Large Language Models and GenAI. - I hold an M.Tech degree in Artificial Intelligence (AI) specialization from the Indian Institute of Science (IISc), Bangalore.- Currently, I work as a Senior Research Engineer at Netradyne, developing AI-driven solutions for transportation safety.- Proficient in Python, SQL, ML / Deep Learning frameworks like PyTorch / TensorFlow, scikit-learn, Langchain, Kubernetes (EKS), Kubeflow, MLflow, MongoDB, AWS, FastAPI, Jenkins, etc.𝗠𝘆 𝗻𝗼𝘁𝗮𝗯𝗹𝗲 𝗿𝗲𝗰𝗲𝗻𝘁 𝘄𝗼𝗿𝗸 (details are in experience section) :* Developed a deep learning model to detect various modes of driver distraction like driver texting on the phone, driver not focusing straight on the road, etc.* Developed a CNN-based model to detect whether the driver is wearing a seatbelt, Created an automated pipeline of stages like data mining, feature preprocessing, model training, validation, etc.* Worked on developing a virtual assistant chatbot using Large Language Models (LLM). Retrieval augmented generation (RAG) and Prompt engineering, enabling dynamic data retrieval from official documents and databases.* Developed a CNN-based multi-head model to detect driver distracted behavior, Created an automated pipeline for end-to-end development lifecycle using docker, Kubernetes, and MLflow.* Improved clustering efficiency of Visual driver login system by replacing Agglomerative clustering algorithm with DBSCAN algorithm.* Created Scalable, Muti-GPU model training infrastructure using MLFlow and Kubernetes. Reduced cost and time for model development.* Experience in building MLOps pipeline/workflows for the entire ML lifecycle. Developed scalable solutions for model training/inference, and optimized resource allocation for high computational loads using orchestration tools like Kubernetes, Kubeflow, MLflow, etc* Developed and maintained pipelines for automated model retraining, ensuring the system can handle continuous learning and updates based on new data