Meghana Ganapa

I’m a Data and AI Analyst with a Master of Data Science from the University of Melbourne, combining analytical depth with practical engineering. I’ve also served as a Data Science Peer-to-Peer Mentor, strengthening my ability to communicate complex concepts clearly and support others through problem-solving.

In my current role as an Agentic AI Engineer, I designed and implemented a Python-based LLM evaluation framework to benchmark model accuracy, latency, escalation rate, and cost-performance tradeoffs. I built automated experiment harnesses with JSONL logging, judge-based escalation, and parser/scoring validation so evaluations stay reproducible and failures are diagnosable. I prioritize questioning model outputs and using AI responsibly to support analysis rather than replace critical thinking.

As an AI/ML Engineer (Contract), I’m contributing to AI-powered decision support solutions for retail category management using LLMs and Graph RAG. I work across backend APIs, data integration, retrieval pipelines, and AI workflows to enable natural language interaction with business data. I aim for explainability, performance, and production readiness—so insights can be trusted and acted on.

My earlier work includes a research internship where I conducted comparative analysis and implemented a CNN in Python, improving tissue classification accuracy by 15%. I’ve also delivered measurable outcomes as an Associate Software Engineer and Software Engineer Intern, including automated workflows that reduced manual effort by 30% and improvements to development velocity and system reliability. Through projects like insurance claims analytics and AI-driven histology extraction, I turn complex data into practical, validated reporting and decision-support outputs.