•Trained a custom NER model based on NLP to extract information from resumes, such as names, organisations, skills, etc.
•Performed matching of the parsed resumes with given Job profiles using Cosine Similarity.
- Data Scientist
Hi, I'm Tanay Sonawane, a graduate of IIT Mandi with international academic and research experience, including a semester abroad and research work in Germany.
Hi, I’m Tanay Sonawane, a graduate of IIT Mandi with international academic and research experience, including a semester abroad and research work in Germany. I am passionate about data science, machine learning, and software development.
I have hands-on experience in backend development, machine learning, cloud deployment, and academic research. My professional journey includes working as a Research Assistant in Germany, as a Software and Research Intern in India, and currently, I am working as a Data Scientist for a U.S.-based client.
•Trained a custom NER model based on NLP to extract information from resumes, such as names, organisations, skills, etc.
•Performed matching of the parsed resumes with given Job profiles using Cosine Similarity.
•Architected and integrated a face recognition model into a Flask-based backend server, enabling automated attendance tracking for 300+
students and efficient data storage using Raspberry Pi.
•Designed and launched a user-friendly web portal for seamless management of attendance records and user data, enhancing
administrative efficiency
Working on XGBoost and Prophet
Leveraged GNPy's Python API to develop a modular simulation system for Free Space Optical (FSO) satellite networks; applied Dijkstra and constraint-based algorithms to compute optimal routes under Gaussian noise constraints. Formulated the overall system design by structuring core components—modulation, transmission, noise modeling, and routing optimal communication paths.
Initiated the development of a deep learning system for traffic violation detection using annotated surveillance video, employing Vision Transformers (ViT) for spatial event localization and BERT for generating contextual, human-readable explanations. Curated a dataset of 1250+ videos through detailed annotation in CVAT, enabling multimodal model training and fine-tuning; implemented temporal flow logic to support transformer attention over sequential frames.
Post a job and invite Tanay, or contact them directly. No hidden commission.