Jyoti Sharma

Senior Associate

STATE STREET•  April 2021 - Present

Project Summary: AI - Reconciliation System. : AI Automating the reconciliation process for Internal Team to improve current conciliation process, This AI engine time have improved the turn around time by 60% Roles and Responsibility: • Analyze and define the requirements and objectives of the AI reconciliation system to ensure alignment with business goals. • Collect, preprocess, and clean data from various sources to ensure high-quality inputs for model training. • Optimize models for better performance and efficiency, including hyperparameter tuning and algorithm refinement. • Deployed the trained models into production environments using tools such as MLflow and Kubeflow. Project Summary: File Anomaly Detection system. : This is Anomaly Detection Engine, We process the file and mark a file as anomaly by detecting there deviation in patter from past data Roles and Responsibility: • Worked on model building for file pattern system, as part of it K-Means Clustering algorithm to automatically determine the SLA of file arrival. • Develop and deploy RESTful APIs for model inference using Flask and FastAPI, facilitating seamless integration with internal systems and applications. • Implement various algorithms and techniques, including supervised and Statistical Learning, • Prepare and present reports and visualizations to communicate findings, progress, and outcomes to stakeholders. • Troubleshoot and resolve any issues related to model performance, data quality, or system integration. Project Summary: KPI Checkpoint: Extracting Data , Processing Documents: Roles and Responsibility: • Used Poppler module to extract tables data and cross validate data across different table extracted. • Developed and selected features for model input, such as embeddings, sentence vectors, or similarity scores. • Created and Deploy NLP models as services or endpoints that can be integrated with backend systems for real-time or batch processing using Flask and Docker. • Implemented and conduct rigorous testing of NLP models to ensure robustness and reliability. • Design and created precise and effective prompts to elicit desired responses from language models for various use cases, such as data extraction, content generation, or decision support using OpenAI API’s.

Senior Member Technical Staff

ADP•  June 2015 - April 2020•  Banglore

Project Summary: Tax and Compliance: Amendment Experience Roles and Responsibility: • Designed, developed, and maintained RESTful APIs using Flask to facilitate the transfer and management of tax amendment data between systems. • Implement database models and integrate with MySQL databases to store and retrieve tax and compliance data. • Implemented security measures to protect sensitive tax data, including encryption, authentication, and authorization mechanisms. • Ensure that the application adheres to relevant tax compliance regulations and standards, including handling and storing data in accordance with legal requirements. • Engaged, Incorporated feedback from clients and stakeholders to improve the functionality and effectiveness of the backend services. Project Summary: Notice Resolution System: The Team worked on resolving the notices sent by agencies to client . • Worked on multi classification model using Ensemble Modelling for classifying the incoming notices redirect to respective team. • This implementation decreases the processing time for notices results in less penalty and improved client experienced by 50% • Developed and train classification models using historical data of notices, fine-tuning hyperparameters to optimize performance and accuracy • Designed and engineered features from raw data, including text extraction, normalization, and the creation of relevant features to improve model accuracy. • Evaluated model performance using metrics such as accuracy, precision, recall, F1 score, and ROC-AUC, and ensure the model meets predefined performance thresholds.