AI Engineer at ELAS, USA

Built a 12-module, multi-tenant operations platform (dispatch, work orders, project management, contracts) with state-machine workflows, RBAC via Keycloak, and RESTful APIs using Gin & GORM, supporting bulk CSV uploads and Swagger documentation. Developed an AI-powered org chart agent using LangChain and Claude with 24+ tool bindings, enabling org-chart creation via text, voice chat, and file uploads; implemented entity merge logic and a LangGraph workflow for extracting, validating, and previewing data from Excel/CSV/PDF files. Designed and implemented an AI-powered Interview Agent using LangChain and Anthropic Claude, automating job postings, dynamic question generation, candidate responses, and scoring through structured JSON-based workflows and RESTful APIs. Deployed microservices to AWS EKS using Docker and GitHub Actions CI/CD, configured Kubernetes with health checks, secrets, and ECR integration, and managed production with kubectl for logging, debugging, and rolling updates.

Data Analyst at University of South Florida, Muma College of Business

Developed Python and ML scripts to automate certificate issuance and detect duplicate records, streamlining administrative processes, providing 100% improvement in accuracy and 40% improvement in total processing time. Maintained a master SQL database for the course, ensuring organized and accurate data management. As a Student Mentor for the Citizen Data Science Certificate program (launched with Tableau), guided undergraduate students through data visualization projects and delivered presentations to explain goals and structure. Supported consumer sentiment analysis and stock market projects, helping students build calculated fields, dashboards, and data-driven narratives.

Data Engineer / Analyst at Accenture India

Designed and implemented an end-to-end automated data pipeline using Automation Anywhere and Python to process large datasets (128 columns, 1000+ rows per file), enabling creation of 100,000 sales and acquisition forms in 48 hours. Built robust ETL logic to ingest, transform, and validate structured datasets before integration with Accenture’s internal CRM application (Manage My Sales). Developed an ETL pipeline using Airflow to automate extraction of economic time-series data from the FRED API (GDP, CPI, etc.). Implemented data cleaning and preprocessing logic, storing intermediate series objects using Python pickle files for caching. Designed automated database load steps using UPSERT procedures in Oracle, ensuring no duplication with dynamic bookmark logic. Monitored the workflow through Airflow DAGs with scheduled refresh and logging. Built a full-stack Flask web application using the MovieLens dataset for personalized movie recommendations via cosine similarity, with Oracle as b