Arun Kumar Parayatham

Dr. Arun Kumar Parayatham 507, Aditya Green Fields, BP Raju Marg, Kondapur Hyderabad, Telangana, India, 500084 --LinkedIn Profile Objective: A highly skilled and accomplished distinguished Data Scientist with over two decades of experience driving data-driven solutions and delivering impactful insights. Adept at leveraging advanced statistical analysis, data science, machine learning, artificial Intelligence, NLP, Generative AI, and, data visualization techniques to solve complex business challenges. Recognized for building innovative data-driven solutions, leading technical teams, and driving successful project execution. Seeking challenging opportunities to contribute expertise and drive transformative change in the field of Data Science. Education: ● Ph.D. in Mathematics, University of Hyderabad, Hyderabad, India, 2000 ● Master of Technology in Computer Science, Jawaharlal Nehru Technological University Hyderabad, India, 2010 ● M.Phil. in Mathematics, University of Hyderabad, Hyderabad, India, 1994 ● M.Sc. in Mathematics, University of Hyderabad, Hyderabad, India, 1992 ● B.Sc. in Mathematics, Physics, Chemistry, Osmania University, Hyderabad, India, 1990 Patents: ● Principal Innovator for 2 algorithms with USPTO assigned patents: ● Method for searching for reliable, significant & relevant patterns (Patent Link) ● Parallel Discretization of Continuous Variables in Supervised or Classified Dataset (Patent Link) ● Patent application submitted for a Distributed Algorithm (Parallel Algorithm) to find reliable, significant & relevant patterns in large datasets. Professional Experience: ​ ● Distinguished Data Scientist, Innominds Software (P) Ltd, Hyderabad, India, 08/2014 to Present ​ ​ ​ ​ ● Resource Principal Mentor for Data Science, Consortium Institutions of Higher Learning (CIHL), MSIT Division of International Institute of Information Technology (IIIT), Hyderabad, India, 07/2008 to Present ● Machine Learning Engineer for Product Development, PurePredictive, Sandy, UT, USA, 05/2013 to 11/2014 ● Freelance Consultant for Data science and other IT related projects 02/2001 to 11/2013 ● National Board of Higher Mathematics (Department of Atomic Energy, Govt of India) Research Associate, University of Hyderabad, Hyderabad, India, 08/2000 to 07/2001 Skills: ● Data Science: Statistical Analysis, Machine Learning, Predictive Analytics, Artificial Intelligence, Decision Science, Big Data, Data Visualization, Optimization Techniques ● Programming: Python, R, Java, C++ ● Platforms: Hadoop, Spark, Azure ML, AWS ML ● Tools: Tableau, Power BI, Jupyter, Git Key Projects: ECG Analysis: ○ ○ ○ ○ ○ ○ ● Spearheaded a groundbreaking project focused on designing and implementing a comprehensive strategy for storing and processing ECG data for analysis purposes. ● Collaborated closely with the Medical Team to create annotations and gain insights into various arrhythmias present in the ECG data. ● Successfully developed models achieving over 95% accuracy in classifying ECG data, enhancing the accuracy of diagnostic analysis. ● Created an end-to-end system integrating data loading, ECG analysis generation, and reporting functionalities. ● Implemented an intuitive UI, allowing ECG technicians to review analysis, perform quality checks, and generate detailed reports efficiently. Facilitated error reporting and annotation capabilities, ensuring continuous improvement in data classification through model retraining. Real Estate Prediction Tool: ​ ​ ​ ​ ​ ​ ​ ● Led the development of a predictive system recommending appropriate geographies for land procurement to build residential properties for profit maximization. ● Utilized advanced models to identify potential markets and clusters for investment over the next 2-3 years. ● Introduced Explainability, providing clear insights into model recommendations and contributing factors, aiding data-driven investment decisions. ● Incorporated What-If scenarios, enabling users to explore investment strategies and their potential outcomes. ● Created comparative summaries for cities and zip codes within a state, projecting future market performance. ● Developed heat maps based on expected returns, guiding investors to potential hotspots and growth opportunities. Developed additional tools to optimize business costs based on land, transportation, skilled labor availability and business expansion opportunities for warehouses, power transformer units ​ iFusion: ​ ​ ​ ​ ​ ​ A platform to provide accelerated Insights and Analytics Solutions at Reduced cost. Secure Data Lake(s), collecting data from heterogeneous, Federated, poly-structured data. Provide access to the data when needed and accelerate the process to build one or more Analytics Solutions at reduced cost and time. Advanced Predictive Platform: ● Developed a platform eliminating inefficient parts of data science, automating model iteration, data preparation, and parameter tuning. ● Introduced automatic discovery of important features and interactions, building ensemble methods for a stable "mega model." ● Enabled efficient and scalable predictive modeling, enhancing the accuracy and efficiency of data analysis. Concept Analytics: ● Created an automated concept retrieval system for large data sets, summarizing analytics and sentiment analysis. ● Provided concept-based search capabilities on vast datasets, facilitating insightful analysis. Automated Job Recruitment System: ​ ​ ● Developed a matching system for resumes and job descriptions, ranking candidates based on thematic alignment. ● Utilized the system in various domains, including automated chatbots. Stock Investor Tool: ​ ​ ● Built a tool providing timely stock advice for optimized gains. ● Analyzed trends, volatility, risk, corporate actions, and market sentiment to advise entry and exit from stocks. Topic Modeling Toolbox: NLP and Generative AI Tools: ● Build tools and platforms to analyze unstructured data and discover semantic knowledge, summary of the key points and actionable items. Build tools to create domain specific knowledge base. ● Created a user-friendly platform for chatbot creation. It allows users ( domain experts) to gather or generate relevant content, then the chatbot provides relevant answers for user queries. Advanced Predictive Platform: This platform does away with the inefficient parts of data science: by doing away with the need to iterate through models, prepare data specifically for each model, impute missing values,and tune parameters. This platform allows to automatically discover important features and interactions, as well as test thousands of transformations of those features, and build ensemble methods. AI pours through thousands of combinations to produce a very stable model. We’ve termed this as a “mega model.” (http://www.purepredictive.com) Concept Analytics: Concept Analytics Does Automated Concept Retrieval. By extracting critical Concepts in Large Data, it summarizes analytics and Sentiment Analysis apart from other analytics. It also provides a concept-based search on large datasets. Topic modeling Toolbox: This toolbox helps users to experiment with several topic modeling techniques and select the best topic model for each data of their text documents. The toolbox also compares all models with performance metrics. Certifications and Achievements: ● National Cadet Corps (NCC) C-Certificate, 1989 ● Captained University of Hyderabad Volleyball Team for Inter University Competitions, 1991, 1992 ● Represented University of Hyderabad Table Tennis Team for Inter University Competitions, 1994 Awards and Fellowships: ● Post-Doctoral Fellowship, National Board of Higher Mathematics (Department of Atomic Energy, Govt of India) Research Associate Fellowship, University Grants Commission, New Delhi. ● National Board of Higher Mathematics (Department of Atomic Energy, Govt of India) Research Award. ● Junior Research Fellowship and National Eligibility for Lectureship, University Grants Commission, New Delhi.