Sonjoy Das Email & Phone Number

Teaneck, New Jersey, Us

Gained as Associate Director, Cognizant Technology Solutions U.S. Corporation• Led multi-disciplinary teams (backend, frontend, database, and business) to develop a Gen AI-based Direct Billing Knowledge Search (DBKS) tool, enabling quicker search capabilities across various document types of different complexities (e.g., PDF, Word, Excel, PPT, HTML, JSON, images, tables) in the insurance domain by leveraging expertise in LLM, Generative AI, RAG, API, ML, and more.• Translated high-level product and business requirements into detailed specifications and designed strategies to build a production-ready API for DBKS.• Collaborated with business stakeholders, technical teams, product owners, and delivery teams to align the Gen AI project with business needs, ensuring successful project completion and smooth delivery.• Actively contributed to designing the evaluation strategy for the DBKS tool.• Applied diverse statistical, machine learning, data science, and Gen AI techniques (e.g., automated evaluation of LLM responses, NLP, OCR) to provide client guidance and accelerate time to proficiency.• Contributed to the pre-production roll-out strategy.• Hands-on experience in end-to-end Gen AI pipeline tasks, including data cleaning, data preprocessing for LLM ingestion, building RAG architecture, prompt engineering, automated evaluation of LLM response, RAGAS, Pre-pilot Streamlit deployment, and developing a production-ready FastAPI application.

Gained as Individual Consultant; as Adjunct Associate professor, Interdisciplinary Unit in Data Science & Analytics, SUNY Buffalo State University; and as Assistant Professor, UB, NY• Significant experience of leading research teams with individuals from diverse technical backgrounds, as well as an ability to function as an independent contributor with a high level of attention to details, as demonstrated by successful completion of 25+ projects with 9+ team members.• Leveraged strong proficiency in programming skills, and solid understanding of the mathematical theory and practical benefits & pitfalls of various Uncertainty Quantification (UQ) and machine learning (ML) methods to extract & manipulate different datasets to build and validate solutions for real-world problems, as evidenced by timely delivery of 5 software to the clients.• Knowledge in data integration, validation and data quality management of ETL and ELT processes (OLTP and OLAP databases), and in Data Lake, Data warehouse, and Data mart.• Experience in creating, configuring, and refactoring tables with advanced SQL queries and Web scraping with BeautifulSoup, Requests, urllib, HTML, and Scrapy.• Implemented various supervised and unsupervised ML techniques, e.g., classification, regression, clustering, bagging, and boosting, in several projects.• Developed and applied neural networks and related concepts, such as, ANN, CNN, RNN, LSTM, Transfer learning, in many different projects.• Managed multiple projects simultaneously by leveraging expertise in probabilistic and statistical methods, such as, random matrix theory, polynomial chaos representations, cross-validation, A/B testing, design of experiments, bootstrap, MCMC, PCA, t-SNE, and copula.• Worked with large datasets of structured and unstructured data (~ 20 million records; ~ 24 GB).• Gathered solid experience with common Unix/Linux tools and version control.• Secured project funding from federal agencies and industries.

Cambridge, Ma, Us

Gained as a Postdoctoral Associate• Demonstrated ability to continuously learn state-of-the-art industry trend to extract, transform and organize data to build and validate predictive models for business problems, resulting in 1 technical presentation on reduced-order model in a conference on applied mathematics and data science.

Los Angeles, Ca, Us

Gained as a Postdoctoral Research Associate• Managed multiple advanced parallel tasks and priorities of different stakeholders, collaborators, and funding agencies for projects in statistical data analysis, vibration analysis, and multi-scale material modeling, resulting in 3 scholarly documents and assisting the supervisor/PI in submission of 1 grant proposal for 1 million USD.

Masters, Applied Mathmatics And Statistics

Phd, Computational Predictive Science

Masters, Probabilistic Mechanics

Be, Civil Engineering