Prompt Engineer

Summary
We are seeking a Senior Prompt Engineer with 5+ years of experience in AI and LLM-driven application development. This role focuses on designing, testing, and optimizing enterprise-grade prompts for GenAI, RAG, and agentic systems. The ideal candidate will have hands-on experience with LLM APIs, vector databases, and multi-agent orchestration frameworks. Strong expertise in prompt engineering, context design, and output control is essential.

Experience with Azure-based AI services, governance, security, and compliance is highly preferred. The role requires balancing accuracy, cost, performance and scalability in production AI systems.

About the Role
In this Role, Your Responsibilities Will Be:

Design, develop, test, and optimize advanced prompt strategies for LLMs (Open AI GPT, Claude, LLaMA, Mistral, Gemini) across enterprise use cases.
Work closely with functional teams to engineer, test, and optimize prompts that translate business requirements into effective LLM-driven application behavior.
Engineer multi-step prompt chains, system prompts, role-based prompts and meta-prompts for complex reasoning, planning, and decision-making tasks.
Develop reusable prompt templates, prompt COT, ToT, Meta Prompting, Conversational Few-Shot and structured flows using frameworks like Lang Chain, Lang Graph, and Prompt Layer.
Develop and maintain prompt orchestration pipelines for RAG, agentic AI, and multi-agent systems using Lang Chain, CrewAI, and MCP-based architectures.
Engineer RAG-optimized prompts integrated with vector databases (Postre

SQL, Redis, FAISS, Pinecone, Chroma) and re-ranking pipelines.
Support agentic and multi-agent frameworks.
Implement function-calling and schema-enforced prompting (JSON/YAML, tool/function APIs).
Design hallucination-resistant, jailbreak-safe prompts with built-in guardrails and filters.
Tune LLM parameters (temperature, top_p, penalties, max tokens) for accuracy–cost–latency balance.
Develop reusable prompt templates, chains, and graphs (Lang Chain, Lang Graph, Prompt Layer).
Perform A/B testing, regression testing, and prompt versioning for continuous optimization.
Optimize token usage and context-window management to reduce inference cost.
Align prompts with fine-tuning / LoRA / RLHF and embedding strategies.
Implement PII/PHI masking, RBAC-aware prompts, and compliance-safe instructions (GDPR, HIPAA, ISO 27001).
Monitor prompt drift, response instability, and output quality in production.
Define and track prompt KPIs (quality, latency, cost, consistency).
Ensure compliance with internal data governance and security policies.

Qualifications
Bachelor's or Master's degree in Computer Science, Data Science, Statistics, Mathematics, or a related field over 5+ years.
Relevant Certifications/trainings on Gen AI, Prompt Engineering.
Strong hands-on experience with LLM prompt engineering in production.
Experience integrating Prompts with RAG systems with vector databases.
Hands-on Prompt Engineer with agentic frameworks and fine tuning.
Familiar with fine-tuning methods.
Strong Python + API integration (FastAPI/Flask).
Knowledge of LLM security, hallucination control, bias mitigation.
Experience/Knowledge in LLMOps/MLOps, MLflow, Lang Smith, W&B, Prompt Layer.
Understanding of PII protection, RBAC, compliance (GDPR/HIPAA/ISO).
Excellent problem-solving, communication, and team collaboration skills.

Preferred Skills
Relevant certifications (Azure AI, Generative AI, LLMOps, Prompt Engineering).

Experience with enterprise copilots, Chatbots, and agentic AI systems.
Hands-on with Prompt Ops / LLMOps governance and lifecycle management.
Expertise in semantic prompting, RAG, and context engineering.

Experience with LLM evaluation tools (RAGAS, Tru Lens, Deep Eval, Prompt Layer).
Strong knowledge of LLM safety, guardrails and hallucination mitigation.
Familiar with hybrid model routing, fallback strategies and Auto Prompting.
Exposure to regulated domains and large-scale AI systems.
Certified in Azure AI Fundamentals (AI-900), Azure AI Engineer Associate (AI-102).

Experience with big data technologies.