Generative AI / LLM Engineer

About the job

We are seeking a highly skilled Generative AI / LLM Engineer with deep hands-on experience in building, fine-tuning, evaluating, and deploying advanced language-model and agentic systems. The ideal candidate has strong technical expertise across LLM training paradigms, retrieval-augmented pipelines, agent frameworks, and AI safety evaluation. Key Responsibilities • Design, implement ... , and optimize LLM fine-tuning pipelines including LoRA, QLoRA, Supervised Fine-Tuning (SFT), and RLHF. • Build and maintain RAG (Retrieval-Augmented Generation) systems using frameworks such as LangChain, LlamaIndex, and custom retrieval layers. • Develop, integrate, and extend applications using Model Context Protocol (MCP). • Architect and deploy agentic workflows using frameworks like OpenAI Swarm, CrewAI, AutoGen, or custom agent systems. • Work with generative AI architectures, including transformer-based and multimodal models. • Implement scalable storage, embedding, and similarity search using vector databases (Pinecone, Weaviate, Milvus, Chroma). • Ensure robust AI safety, including red-teaming, adversarial testing, and evaluation of model behavior. • Collaborate with cross-functional teams to deliver end-to-end AI-driven features and products. • Monitor performance, reliability, and quality of deployed AI systems, optimising continuously. Required Skills & Experience • Strong, hands-on experience with LLM fine-tuning : LoRA, QLoRA, SFT, RLHF. • Deep expertise with RAG frameworks and retrieval pipelines (LangChain, LlamaIndex, custom retrieval layers). • Practical experience with MCP (Model Context Protocol) for tool integration and orchestration. • Proven work with agent frameworks (OpenAI Swarm, CrewAI, AutoGen, or custom agent systems). • Solid understanding of transformer architectures, generative AI models, and multimodal systems. • Proficiency with vector DBs : Pinecone, Weaviate, Milvus, Chroma. • Strong grounding in AI safety, red-teaming strategies, evaluation methodologies, and risk assessment. • Experience with Python, distributed systems, and MLOps tooling is a plus.

Requirements

• LLM fine-tuning
• RAG frameworks
• MCP
• agent frameworks
• transformer architectures
• vector databases
• AI safety
• Python

Preferred Technologies

• LLM fine-tuning
• RAG frameworks
• MCP
• agent frameworks
• transformer architectures
• vector databases
• AI safety
• Python

About the company

TAC Security is focused on developing advanced language-model and agentic systems and offers a highly skilled environment for AI professionals.