Yadniki Sonawani

I am a results-driven Gen AI engineer with hands-on experience designing and deploying next-generation AI solutions that bridge business needs with research. I combine data science rigor with creative problem-solving to deliver measurable impact.

At e-Zest, I designed and optimized RAG pipelines, built LLM-powered chatbots, and implemented computer vision workflows for automated image/video analysis. My work focused on integrating FAISS vector databases, Hugging Face embeddings, LangChain, FastAPI, and Streamlit to improve response accuracy and context retention.

Previously, I worked as a data analyst intern across gamification and automotive domains, building interactive dashboards with Power BI, Python, and SQL, applying A/B testing and statistical analysis to improve user retention and reporting efficiency. I reduced development and reporting times through practical, production-oriented solutions.

I enjoy turning complex problems into scalable PoCs and MVPs, collaborating with cross-functional teams, and shipping AI-driven products that deliver business value. I prioritize clean deployment patterns, user-centered interfaces, and reproducible pipelines.