Research Assistant Data Science

Mitesh Jalan

H +1-631-***-**** # ad5n9n@r.postjobfree.com ï linkedin.com/in/MiteshJalan/

§ github.com/MiteshJalan miteshjalan.github.io Experience

Research Assistant Aug 2023 – Present

State University of New York at Stony Brook - Advisor: Dr. Fusheng Wang Stony Brook, USA

• Architectured a Multimodal Transformer Model combining clinical notes and structured patient data to predict mortality risk, harnessing the MIMIC-III dataset with 100K+ clinical records.

• Performed comparative analysis of the model resulting in this model performing better than other models.

• Developed a clinically relevant patient summarization model by utilizing 5+ pre-trained LLMs (2 trained on medical data), tailored specifically for healthcare applications Data Science Intern May 2022 - Dec 2022

Sony Research India Bangalore, India

• Improved session-based content recommendation system that’s benchmarked on 6000 users data.

• Utilized BERT Embed dings, with Hierarchical Recurrent Neural Networks with 2 GRUs, leveraging segmented user interactions for session-based recommendations

• Achieved a 6% increase in Recall@20 and 3% improvement in MRR@20. Data Associate Intern Mar 2022 - Apr 2022

Unscript. A.I. PVT LTD Bangalore, India

• Delivered tools to automate audio noise suppression, saving 200+ hours monthly in subtitle generation.

• Automated audio-to-subtitle workflow reducing generation time from 1 hour to 5 minutes per file.

• Made company monetary savings of more than 15K INR per month. Software Developer Intern May 2021 - Aug 2021

Pocket FM Bangalore, India

• Enhanced CMS Dashboard and API features, which had a daily utilization rate of 7% in the first week of launch.

• Implemented podcast tagging system to optimize searchability and ranking, resulting in improved discoverability and audience engagement.

Education

State University of New York at Stony Brook Stony Brook, USA Master of Science in Computer Science Graduate Teaching Assistant Jan 2023 - Dec 2024(Exp.) Hindustan University Chennai,India

Bachelor of Technology in Computer Science Jul 2018 - May 2022 Projects

Ragwarts RAG LLM Model NLP, Sentence transformer, VectorDB, LLM, Pytorch, LangChain GitHub

• Implemented RAG model for answering questions on Harry Potter Universe with 2M+ tokens.

• Conducted Fine-tuning and Instruction tuning on 2 LLM models to optimize for domain-specific goals, resulting in a leading model achieving a Rough-L score improvement of 65%.

• Employing LoRA and Quantization techniques to minimize hardware overhead reducing training time by 60%. Story Recommendation Model Machine Learning, Collaborative filtering, Matrix Factorization GitHub

• Used open-source data to construct a collaborative filtering model on a Dataset of 1000 users.

• Employing matrix factorization techniques such as SVD, SVD++, and neural matrix factorization. AI Traffic Police Computer Vision, TensorFlow, Easy-OCR, YOLOv4, Object detection GitHub

• Built a real-time license plate detection and recognition model with TensorFlow and Easy-OCR, improving accuracy and efficiency in identifying license plate numbers. Skills and Interests

• Technical Skills: Python, JavaScript, Yaml, Flask, Django, React, PyTorch, TensorFlow, Firebase, MySQL, MongoDB, AWS (EC2, S3, Transcriber), Docker, Git, Linux, Postman, Hadoop, Spark, Tableau, Nginx, RabbitMQ, Kubernetes.

• Interests: Machine Language Understanding, Cloud Technologies (AWS), Infrastructure, Micro-services, Web Development, DevOps, CI/CD

Contact this candidate