Prompt Engineer

Work on frontier Generative AI systems used at massive global scale Hands-on prompt engineering across Gemini and other frontier LLMs High-impact role driving 95%+ alignment with human reviewer performance Collaborative, experimentation-driven environment with client calibration Hyderabad-based role with clear technical growth pathways Hands-on experience with Gemini, OpenAI, or Anthropic APIs Strong expertise in Prompt Engineering (few-shot, system prompts, structured outputs, context management) Strong Python development skills (not notebook-only usage), with hands-on prompt engineering design, deployment, and debugging Experience debugging, testing, and improving LLM applications Strong analytical thinking and an experimental mindset Ability to write clear technical documentation and RCA reports Comfort working in ambiguous, evaluation-driven environments 5 to 5+ years of experience; based in Hyderabad Good to have: Agentic AI, ReAct, and Chain-of-Thought prompting Good to have: RAG systems and evaluation datasets Good to have: LangChain, LangGraph, or LlamaIndex Good to have: REST APIs, Postman, and JSON Good to have: Trust & Safety, Content Moderation, or Ad evaluation experience Tune and test Gemini and LLM-based systems in production workflows Analyze overnight evaluation outputs and identify model failures Classify issues such as hallucinations, context misses, and boundary errors Design and iterate prompts using hypothesis-driven experimentation Improve prompts using few-shot examples, constraints, and RAG updates Run regression testing using golden datasets before deployment Deploy prompt changes and monitor system performance Write RCA reports and hypothesis documentation, and participate in client calibration sessions Work with agentic AI and tool-calling workflows Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.