OverviewAI Engineer - Defence RAG Systems ( Security Clearance Essential ) On Site 2 X Days a week Plymouth - willingness to undergo DV if requiredRole OverviewDefence client requires an SC Cleared AI Engineer to build fully on-premises RAG systems using open-source technologies. You'll develop classified AI capabilities on air-gapped infrastructure with zero external dependencies.Key ResponsibilitiesBuild end-to-end RAG pipelines on isolated defence networks using open-source LLMs (Llama 3, Mistral, Qwen)Deploy local vector stores (Chroma, FAISS, Milvus) with sensitive document ingestion pipelinesHost and optimise LLMs using vLLM/TGI on local GPU clusters without internet connectivityImplement agent orchestration using LangChain/LangGraph in completely offline environmentsDesign secure document processing for classified materials with appropriate data sanitisationBuild monitoring and evaluation systems that operate within air-gapped infrastructureEssential RequirementsActive SC Clearance (non-negotiable) - willingness to undergo DV if requiredDemonstrable experience deploying open-source LLMs (Llama, Mistral, Falcon) on-premisesExpertise with local vector databases (Chroma, FAISS, Weaviate) in offline deploymentsStrong vLLM/Text Generation Inference experience for high-throughput model servingProven ability to work on air-gapped systems with no external package repositoriesExperience with GPU orchestration (NVIDIA A100/H100) and CUDA optimisationPython expertise with offline dependency management and local package mirrorsTechnical Stack (All On-Premises)Models: Llama 3, Mistral, Qwen (locally hosted)Vector Stores: Chroma, FAISS, MilvusOrchestration: LangChain, LangGraph for agentsHosting: vLLM, TGI, Ollama on bare metal/private cloudInfrastructure: Air-gapped Kubernetes, local container registriesDesirable SkillsExperience with defence/government IT security protocolsKnowledge of CIS benchmarks and NCSC guidelinesFamiliarity with cross-domain solutions and data diodesUnderstanding of classification marking and handling procedures #J-18808-Ljbffr