Harish Krishna
AI Security Engineer | LLM Guardrails & Red-Teaming | Agent Safety | LLMOps
Hire this person
Send a job offer directly to this candidate
About
AI/ML Engineer with 2+ years building production LLM and agentic systems end to end — retrieval-augmented generation, knowledge-graph construction and multi-agent orchestration. Builder of advanced RAG engines (hybrid retrieval, re-ranking, HyDE, CRAG) and text-to-knowledge-graph tooling, with an open-source, PyPI-published AI guardrails gateway. Deployed enterprise AI serving thousands of daily interactions. Strong in LangGraph, MCP, evaluation and LLMOps.
Experience
Associate Software Engineer – AI at AI Ensured (TestAIng.com) (2024-09 – Present)
- Built end-to-end agent evaluation frameworks (RAGAS, LLM-as-Judge, PromptFoo) and led AI security testing — jailbreak resistance, prompt-injection detection, adversarial robustness — for enterprise LLM deployments.
- Fine-tuned LLMs (LoRA, SFT) for domain adaptation; built a bias-detection suite adopted across enterprise client engagements; promoted within 8 months.
- Deployed AI products for 3+ enterprise clients in high-availability production serving thousands of daily interactions; built PowerBI/Tableau dashboards for performance and fairness reporting.
Machine Learning Engineer (Intern) at AI Ensured (TestAIng.com) (2023-12 – 2024-09)
- Architected a multi-agent RAG chatbot processing 1,000+ daily queries at 88% accuracy using semantic chunking, re-ranking, and A/B-tested LLM selection.
- Shipped 15+ production features including hallucination monitoring and explainability tools; contributed forecasting and churn-prediction models used for resource allocation and retention.
Education
B.E. in Electrical & Electronics Engineering – Anna University, Chennai (2013 – 2017)
