saihemanth mandala

I’m a Software Engineer with hands-on experience in applied AI systems, data engineering, and backend development. I focus on practical problem-solving, system reliability, and outcomes you can measure.

At DHIRA, I refactored data profiling workflows, improving execution performance by ~50%. I built a context-aware textbook Q&A using Retrieval-Augmented Generation (RAG), plus an AI-based JD–Resume matching system with a Gradio UI for document upload and results visualization.

I also developed a POC Change Data Capture replication pipeline from MariaDB to PostgreSQL using Debezium, Kafka, and Spark Streaming. To strengthen data governance, I implemented ETL transformations and data-quality checks (including data masking and normalization) and registered Apache Atlas entities to improve lineage visibility and reduce data issue investigation by ~30–40% across workflows.

Beyond that, I evaluated Delta Lake by building a pipeline using Spark, Hive, and HDFS, and I implemented an NLP-driven SQL generation workflow to turn natural language into executable queries. Earlier, as a Data Science and Machine Learning Intern, I worked on data analysis and foundational ML tasks like linear regression.