Ruthu Rajendra

About

Built and productionized an LLM-powered customer support system using a RAG architecture with structured document ingestion, embedding retrieval, and prompt templates for product-specific grounding. Experienced diagnosing live agent/workflow failures (e.g., retrieval regressions after new docs) by refactoring ingestion/chunking and adding grounding constraints plus evaluation benchmarks. Also supports go-to-market by joining discovery calls, shaping MVP workflows into demos/prototypes, and creating post-launch documentation to drive adoption.

Experience

Education

Key Strengths

• Takes LLM systems from prototype to production with accuracy, explainability, and safety focus
• Designs and improves RAG systems (embeddings + prompt templates) to increase relevance and grounding
• System-level real-time troubleshooting: isolates failures via impact assessment, log inspection, and targeted fixes
• Root-cause analysis of retrieval regressions (e.g., inconsistent chunking/noisy data) and stabilizes accuracy via ingestion refactors
• Builds evaluation/benchmarking “decks” to catch issues pre-deployment
• Adapts technical demos/workshops for mixed audiences; dives into architecture/APIs and implementation details when valuable
• Partners with sales/customer-facing teams by translating discovery into MVP workflows, prototypes, and rollout documentation