Sheryl Peter Email and Phone Number

Developed a predictive analytics tool in Python to optimize inventory management within the pharmaceutical supply chain, enhancing operational efficiency.Employed machine learning algorithms using Pandas and Scikit-learn to analyze historical sales data, enabling accurate forecasting of inventory needs based on trends and patterns.Integrated the predictive analytics tool with existing ERP systems through RESTful APIs, ensuring seamless data exchange and enhancing overall system interoperability.Achieved a 20% reduction in excess inventory and a 15% decrease in stockouts over six months, significantly improving supply chain efficiency and reducing operational costs.Conducted thorough data validation and cleaning processes to ensure the accuracy and reliability of input data for machine learning models, improving forecast reliability.Implemented user authentication and authorization mechanisms using Django's built-in features, resulting in a 40% reduction in security vulnerabilities and ensuring secure access to the application while protecting sensitive inventory data.Designed a data model in MongoDB that effectively represented complex relationships between inventory items, suppliers, and historical sales data, enabling efficient querying and reporting. Utilized Amazon S3 for storing large datasets, achieving a 30% reduction in data storage costs compared to previous solutions, while ensuring easy access for analysis and enabling scalable storage solutions for historical inventory data. Participated in Agile ceremonies, including daily stand-ups, sprint planning, and retrospectives, which enhanced project management and improved delivery timelines by 15% through effective collaboration.

Developed a LLM which could efficiently switch between the two microservices which are NL to SQL and NL to REST API.An end-to-end approach was devised for interacting with structured databases in natural language with metadata extraction and OpenAI’s LLM, offering a system that simplifies user database interactions by eliminating the need to query against the database programmatically. The API provided metadata extraction, SQL generation, and result summarization capabilities that serve as a foundation for constructing SQL queries that reflect the user’s intention.In NL to REST API, the approach combines several components, from OpenAPI specification to function creation, K-means clustering, and semantic embedding, resulting in efficient communication between human language and API endpoints. The solution leverages NLP techniques and OpenAI function calling capabilities.This project was containerized and was successfully deployed on Azure.The LLM successfully provided accurate answers, although it showed a delay in responding to complex questions, and slight hallucinations were noted. Despite these observations, the overall performance of the LLM was satisfactory. Subsequent efforts were made to enhance its capabilities and it was tailored for client-facing applications.Built a chatbot that could integrate this LLM with MS Teams to make it available for all the employees. Worked on Power Apps and built the chatbot using Power Virtual Agent (PVA).

Implemented a scalable e-commerce platform utilizing Python with Django for the back end and React for the front end, boosting user engagement by 30% through an intuitive UI/UX design that streamlined the shopping experience.Optimized PostgreSQL database queries through indexing and query restructuring, resulting in a 40% reduction in data retrieval times and significantly enhancing application performance during peak usage.Created interactive visualizations using D3.js and Chart.js, enhancing user understanding of complex data trends by 45%, allowing for quicker actionable insights and better strategic decision-making.Designed a microservices architecture to enhance application scalability, accommodating a 60% increase in concurrent users without compromising performance.Made a Flask-based admin panel for content management, allowing easy updates and modifications to product listings, enhancing operational efficiency by 30%.Automated data reporting processes through Python scripts, reducing manual reporting efforts by 70%, allowing data analysts to focus on insights instead of data preparation, thereby boosting productivity.Developed data aggregation pipelines using Apache Kafka, processing over 10,000 data points per second, which enabled clients to gain immediate insights and improved operational efficiency.Created RESTful APIs to facilitate seamless data exchange between the front end and back end, enhancing overall system responsiveness by 25%, leading to faster load times and improved user satisfaction.Conducted thorough unit and integration testing using Pytest and Selenium, achieving a 99% defect-free release cycle, which ensured high product reliability and minimized post-launch bugs.