Hi, I'm Purva Desai

Prepared clear and concise reports, presentations, user manuals, and technical documents related to ETL processes, SQL queries, MVC frameworks, Python scripts, Flask applications, database management, building data pipelines, and automating data flows. These communication efforts ensured that both technical and non-technical stakeholders were informed about the status of our projects. Worked closely with developers and engineers to interpret large datasets using SQL and Python, deriving meaningful insights to support the projects and initiatives.

• Managed and supported enterprise-level applications, ensuring optimal performance and minimal downtime.
• Developed and maintained ETL processes to extract, transform, and load data from various sources into databases, enhancing data integration and accessibility.
• Utilized Power Automate to streamline workflows and automate repetitive tasks, improving operational efficiency.
• Leveraged Python for scripting and automation, enabling advanced data analysis and process automation.
• Collaborated with cross-functional teams to gather requirements, design solutions, and implement new features.
• Provided training and support to end-users, enhancing their ability to effectively utilize IT applications.
• Led and contributed to key projects, including "Fusion BI" and "WIP Report," delivering impactful business intelligence solutions and work-in-progress reporting tools.

Interacted with clients, gathered data of the current system, and provided either new design or solution from the current system after doing proper data analysis and based on client requirements. Also responsible for creating UI/Wireframes using Microsoft Visio. Initiated Face Recognition related projects and worked in cross-functional teams. Trained Face Recognition algorithms by accumulating user data (face minutiae), then implemented clustering to identify users from their faces. Was accountable for inter-team conflict resolution. Accountable for making the algorithm adaptive so that it learns by itself, changing the face of the user and identifying the user correctly after years. Was rewarded with the “Made a Difference” member award for the same. Provided a complete data analysis report on client demand, in which the client wanted data of all its users. Utilized machine learning techniques to train models, derive insights, and ensure accurate predictions.

• Led Award-Winning Face Recognition Project: Spearheaded a cross-functional team of 9 members to develop a face recognition system achieving a 93% accuracy rate in user identification. Trained the algorithm using a dataset of 1000 user faces. Implemented clustering techniques and received the "Made a Difference" award for this project's impact.
• Increased Client Satisfaction through in-depth interviews with 20+ clients to gather data and deliver 15+ custom UI/UX design solutions (Microsoft Visio) that addressed their specific needs.
• Streamlined Data Management and Processing: Reduced processing time by 20% by managing large datasets (50 gigabytes/entries) using Excel and a MySQL database, formulated business logic to optimize processes. Documented workflows for improved transparency.
• Delivered Data-Driven Insights: Provided comprehensive data analysis reports in response to client requests. Analyzed user details, credentials, and entry/exit times for 5 clients, generating actionable insights to improve operations.

Provided a complete data analysis report on client demand, in which the client wanted data of all its users. Utilized machine learning techniques to train models, derive insights, and ensure accurate predictions. Installed complete IoT hardware in the computer laboratory for a day to collect data. The occupancy of chairs and tables was diagnosed with 95% accuracy.

• Detected occupancy of chairs and desks maneuvering accelerometer where Raspberry Pi was deployed as platform.
• Collected 3D coordinate data(x,y and z) using an accelerometer which was converted to a tabular format (csv format) followed by an algorithm to operate data for detection of presence of anyone on chairs or tables.
• Installed complete hardware in the computer laboratory for a day for data collection and occupancy of chairs and tables was diagnosed for which 95% of accuracy was achieved.

How a Smart Face Recognition System Made High-Tech Security Accessible to All

The Daily Scanner