vandana basrani

Core engineer on a 3-person team driving backend development, system architecture, and cloud infrastructure. Built a fuzzy semantic search engine and an internal Quality Control portal, cutting latency by 70% and ops effort by 85%.

Built high-performance reactive batch pipelines using Spring Boot for large-scale XML ingestion, improving data processing efficiency by 60%. Orchestrated event-driven batch jobs with IBM MQ on Kubernetes/GKE and automated schema migrations with Liquibase.

Enhanced model deployment and API integration with LLMs, Hugging Face, FastAPI, and advanced databases. Implemented Mistral-7B and Llama-2-7B instruct models, improving inference speed by 35% and model efficiency by 40%.

Built a lightweight deep learning model to predict human emotions from images using facial landmarks. Extracted 68 facial landmarks with dlib and trained a DNN on FER-2013, improving accuracy by 6%.

Grade: 7.7/10

Bachelor of Technology in Electrical Engineering from the Indian Institute of Technology Kanpur, with a CPI of 7.7/10.