AI Prompt Engineering Consultant
Technically Sharp | Systems-Minded | GenAI Focused
Design, optimize, and operationalize prompt-driven and agentic AI systems. Architect LLM-powered workflows that connect people, data, and intelligent systems in high-impact, production-ready ways.
THE ROLE
Prompting, Reasoning & Agentic Systems
* Design, test, and optimize prompts for leading frontier models (GPT-4.x/5, Claude 3+, Gemini, LLaMA, DeepSeek, and emerging open-weight models).
* Apply advanced prompting and reasoning techniques, including:
* Chain-of-Thought, ReAct, Tree-of-Thoughts, Graph-of-Thoughts, Program-of-Thoughts
* Self-reflection and critique loops
* Debate prompting and multi-agent collaboration
* Architect agentic workflows using frameworks such as AutoGen, CrewAI, LangGraph, and custom orchestration layers.
* Build systems with tool calling, long-term and short-term memory, retrieval pipelines, and structured reasoning constraints.
GenAI Application Engineering
* Integrate LLMs into real-world applications using LangChain, LlamaIndex, Haystack, AutoGen, and OpenAI Assistant / Responses API patterns.
* Design and implement high-performance Retrieval-Augmented Generation (RAG) pipelines, including:
* Hybrid (keyword + vector) search
* Reranking and embedding optimization
* Chunking and document preprocessing strategies
* Evaluation and regression testing harnesses
* Develop APIs, microservices, and serverless GenAI workflows for scalable, secure deployment.
ML / LLM Engineering & LLMOps
* Work across AI/ML platforms such as Azure ML, AWS SageMaker, Vertex AI, Databricks, Modal, and Fly.io.
* Deploy and manage vector databases and embedding stores, including Pinecone, Weaviate, Milvus, FAISS, ChromaDB, and pgVector.
* Implement LLMOps / PromptOps practices using tools such as:
* Weights & Biases, MLflow, LangSmith, LangFuse, PromptLayer, Humanloop, Helicone, Arize Phoenix
* Benchmark, evaluate, and monitor LLM systems using RAGAS, DeepEval, custom eval suites, and human-in-the-loop review.
* Leverage AI-native developer tools (GitHub Copilot, Cursor, Codeium, Aider, Windsurf) to accelerate iteration and experimentation.
Deployment, Performance & Infrastructure
* Containerize and deploy GenAI workloads using Docker, Kubernetes, KNative, and managed inference endpoints.
* Optimize system performance with:
* Caching, batching, routing, and fallback strategies
* Quantization and distillation for efficient inference
* Cost, latency, and reliability optimization
* Design resilient, observable GenAI systems suitable for production environments.
EXPERIENCE
* Strong Python engineering skills with hands-on experience across the modern GenAI ecosystem.
* Deep understanding of LLM behavior, prompt optimization, embeddings, retrieval strategies, and data preparation workflows.
* Practical experience with vector databases and semantic search systems.
* Comfortable working in Linux environments with Bash/PowerShell, containers, and cloud infrastructure.
* Strong communication skills, creativity, and a systems-thinking mindset.
* Curious, adaptable, and motivated to stay ahead of rapid advances in GenAI and AI-native software development.
BENEFICIAL
* Experience with PromptOps, LLM observability, and evaluation tooling.
* Understanding of Responsible AI, safety, bias mitigation, governance, and compliance frameworks.
* Background in Computer Science, AI/ML, Engineering, or a related technical discipline.
* Experience deploying, fine-tuning, or serving open-source LLMs in production.
Staffworx is a UK-based Talent & Recruiting Partner supporting organisations across Digital Commerce, Software Engineering, and Value-Add Consulting sectors throughout the UK & EMEA.