Ravindra Pandey

I’m a Generative AI and large-scale data systems engineer with over 2+ years of experience productionizing state-of-the-art architectures. I specialize in privacy-centric RAG systems and multi-agent orchestration frameworks that automate complex research and synthesis tasks.

At Inxite Out, I build and optimize LLM-driven analytics and automation systems using RAG pipelines, Snowflake, Databricks, and Python. I engineered a Critique-Iterative multi-agent synthesis framework that achieved a 90% reduction in manual effort, and migrated preprocessing workflows from Pandas scripts to Databricks, boosting pipeline performance by 40%.

I’ve also developed fully offline, low-latency RAG infrastructure—using Llama 3.2 (8B), llama.cpp, Mem0, FAISS, and GGUF quantization—to power privacy-first research over historical archives. Beyond GenAI, I analyze 100B+ record datasets in Snowflake and optimized TrueSkill/Elo-based matchmaking analytics in AWS SageMaker, reducing manual effort by 95%.