Avaneesh Khandekar Email & Phone Number

United States

Jacksonville, Florida, United States

◦ Developing a custom Retrieval Augmented Generation (RAG) Q&A chatbot to provide insights on the organization’sfinancial and budget data, incorporating NL2SQL to reduce time spent on financial data analysis◦ Created internal apps and automated business processes with PowerApps and PowerAutomate

Pune, Maharashtra, India

• Worked as a full stack cloud software developer for a leading investment advisor and asset management company in the BFSI unit.• Implemented AWS-based event notification system using Lambdas, SQS queues, and SNS notifiers to enhance client appointment scheduling. Resulted in reduced missed appointments by sending real-time and reminder SMS notifications to clients.• Built AWS ETL application using DataSync, Storage Gateways, Lambdas, and SQS queues. Extracted data from mainframe server and seamlessly transferred it to MS Dynamics CRM in a modern, accessible format.• Set up AWS Multi-Region and Out-Of-Region cloud contingency mechanisms for microservices and web apps. Ensured disaster recovery and resiliency, eliminating downtime when switching regions during outages.• Implemented REST web services (Java & Spring) and micro-UIs (Angular & Node.js) on AWS for a modernization project, following the microservices architecture. This approach enhanced speed, usability, and cost reduction compared to the legacy system, leveraging AWS services.• Developed a PII data masking library for automatic JSON log masking in web services (REST API). This add-on, known for its simplicity and enhanced security, has become the default organization-wide solution for the client.• Conducted training sessions for new hires and trainees on Java, AWS, Angular, Agile, and Kanban technologies and business practices.• Worked on migrating on-premise applications to AWS ECS.

Pune, India

• Project intern for Capstone undergrad proect (BE)• Created a fuse box assembly defect detection system that identifies incorrect fuse placements based on colors. Eliminated the need for human inspection, resulting in reduced operating expenses and labor-intensive quality control.• Utilized Image Processing and a two-step Neural Network with TensorFlow and Faster RCNN models to detect improper fuse box configurations in images, videos, and live feeds. Enhanced model efficiency and performance through dataset creation, augmentation, and fine-tuning techniques.• Developed a desktop application with GUI interface, packaging the model for user-friendly usage. Added business requirements such as generating PDF reports with detailed results for each fuse box, charts, total error rate, and most frequent errors for auditing.

Master Of Science - Ms, Computer Science

Bachelor Of Engineering - Be, Computer Science, 8.78

Computer Science, 69.85%

Higher Secondary Certificate (XII th)

High School, 93.20%

Secondary School Certificate (Xth)