Sai Gautham Ghanta

About

AI/full-stack engineer who has built both a real-time crypto sentiment platform from scratch and production enterprise RAG search systems at Kore.ai. Stands out for combining strong systems engineering with practical LLM evaluation, retrieval tuning, and careful human-in-the-loop design for high-risk network automation use cases with Cisco.

Key Strengths

• End-to-end full-stack ownership of AI products
• Strong RAG and search relevance engineering
• Production LLM system design with grounding and fallback handling
• Hybrid retrieval optimization using BM25, dense vectors, and rerankers
• Applied NLP model validation without formal labeled datasets
• Real-time data platform architecture with sub-1-second p95 latency
• User-focused AI feature iteration based on feedback loops
• Evaluation-driven development with curated regression test sets
• Production monitoring and regression detection using New Relic dashboards
• Domain adaptation of embeddings and retrieval for enterprise knowledge bases
• Thoughtful human-in-the-loop decision making for high-risk automation
• Early-stage technical leadership in ambiguous environments
• Cross-functional alignment around concrete user pain points
• Quantified impact: 90% answer quality improvement, 25% top-k hit rate lift, 30% less irrelevant context, 20% fewer no-result queries

Experience

Education

Skills

Languages