GenAI based conversational data insight tool: Led and won a GenAI based hackathon project by developing an innovative conversational service capable of extracting and interpreting insights from numeric datasets. Successfully transitioned the project from concept to production, enhancing data analysis efficiency and user experience.

1. Utilized advanced quantitative methods and machine learning techniques to develop and implement a sophisticated revenue forecasting model. This model employed trend analysis and statistical forecasting to accurately predict future business outcomes for customer. The successful application of these techniques led to a significant 15% improvement in revenue forecasting accuracy. Additionally, the strategic insights generated by the model contributed to a 10% increase in revenue growth in the upcoming fiscal year, demonstrating the model’s impact on data-driven decision making and business strategy enhancement.2. I led a collaborative initiative with key business stakeholders to understand their goals and challenges. This resulted in the design and deployment of an innovative unsupervised machine learning model. My approach focused on quantifying media partner profiles through a unique ‘Trust Score’ metric. This metric not only provided actionable insights but also directly contributed to a 10% increase in marketer engagement with our product, bolstering its trustworthiness and reliability in the market.3. Led the development and deployment of a text classification system for persona classification, utilizing Sentence BERT embedding to achieve an impressive accuracy rate of 96%, significantly optimizing precision marketing endeavors.

1. Developed and deployed a novel deep learning model for identifying statistical mediators in a high dimensional clinical dataset. The model results in identifying distributed, network-level neural pathways that mediate the relationship between stimulus intensity and reported pain.2. Led the development of an advanced statistical model utilizing machine learning techniques to analyze the impact of interventions on health outcomes. Demonstrated comprehensive engagement management skills by overseeing project planning, budgeting, and economics, successfully securing a $25,000 Oracle research grant. This achievement not only highlights my leadership capabilities but also showcases my ability to effectively manage risk, define deliverables, and persuade top-level management through compelling grant proposal writing.3. Authored a comprehensive book chapter on “Machine learning applications in medical image processing” contributing insights on cutting-edge algorithms and methodologies used in real-world business and academic scenarios.4. Developed a groundbreaking new field at the intersection of AI and body metabolism, leading to the publication of multiple papers in top-tier scientific journals, advancing our understanding of the complex relationship between artificial intelligence and physiological processes.

1. Developed DeepLabCut 2.0 a deep learning with Tensorflow based algorithm for markerless pose estimation to a self contained pip installable python package. As a result, the package has over 650,000 downloads and is used for 3D markerless pose estimation across species and behaviors.2. Published the package in Nature protocols, a renowned academic journal, with the article ranking among the top 3 most downloaded papers in the field of Neuroscience for the year 2019.3. Seamlessly integrated OpenCV with DeepLabCut to process real-time video feeds of Cheetah hunting in the wild, enabling the generation of full-body 3D reconstructions. The developed tool revolutionized the study of Cheetah hunting behavior, providing scientists with unprecedented insights into their natural hunting patterns.4. Led the design and implementation of a robust end-to-end data curation and ingestion pipeline using Datajoint, streamlining the process of data management, integration, and retrieval. This solution facilitated seamless collaboration across multi-disciplinary teams, ensuring data integrity and accelerating insights extraction for critical research projects.

1. Deployed statistical model to identify the neural correlate of inattention in clinical population using high dimensional neuroimaging dataset. The model resulted in identifying the associations between clinical traits of inattention and white matter organization of the brain.2. Collaborated effectively with across 19 institutions across different time zones, showcasing strong communication skills, to play a pivotal role in supporting large-scale data collection and organization for ABIDE 2. This endeavor was critical in promoting innovative approach on understanding the brain connectome in Autism Spectrum Disorder (ASD). Additionally, conducted a comprehensive data quality check to ensure the reliability and accuracy of the collected data.3. Spearheaded the development of an in-house, comprehensive end-to-end data processing pipeline for daily incoming data, significantly enhancing data management efficiency and streamlining data processing workflows.4. Mentored and nurtured junior data scientists, empowering them with invaluable guidance and fostering their professional growth. Utilized a hands-on approach to provide personalized coaching, leading to substantial improvements in their technical skills, problem-solving abilities, and overall performance.

1.Development of imaging platform to record the activities of animals/insects.2.Algorithm development for quantification of socialbehavior of animals.

1. Built a machine learning model for tracking multiple objects: Developed a robust computer vision-based machine learning algorithm for real-time tracking of multiple objects, enhancing accuracy and performance in dynamic environments.2. Multi-site collaboration: Collaborated with cross-functional teams to integrate the tracking algorithm into existing systems, ensuring seamless functionality and improved overall system efficiency.3. Effective verbal communication: Presented findings and technical details to non-technical audiences, including senior business executives and managers, effectively communicating the algorithm’s benefits and impact on business operations.

Computational Neuroscience and Medical Image Processing

Worked on developing "A method for Volume Reconstruction by Inverse Interpolation and Automated Brain Landmark Detection"Worked as a Visiting Scholar under Dr. Hans J. Johnson Professor, Department Of Psychiatry, University Of Iowa, Iowa City, IA, USA to complete my M.Tech thesis. The project is based on removal of Interleaved Artifacts in MRI by registration process using Computational Morphological Tool Kit (CMTK). The detection of Brain Landmarks is based on Principal Component Analysis and Statistical Shape Modeling.