AI/ML staffing in New York and North Carolina
Triangle Workforce recruits machine learning and AI engineers for companies building real systems - from LLM products and retrieval pipelines to training infrastructure and inference at scale.
Roles we recruit
- Applied AI and LLM engineers building products on GPT-4-class APIs, open-weight models (Llama, Mistral, Qwen) and fine-tuned in-house models.
- ML engineers and applied scientists: PyTorch, JAX and TensorFlow, transformers, recommendation and ranking systems, computer vision and speech.
- MLOps and model platform engineers: training pipelines, model registries, CI/CD for models, evaluation harnesses and observability.
- Infrastructure engineers for AI: distributed training (FSDP, DeepSpeed, Megatron-LM), GPU clusters, CUDA, NCCL, Kubernetes and Ray.
- Inference and serving engineers: vLLM, TensorRT-LLM, quantization, KV-cache management, latency and cost optimization.
- Data engineers for AI systems: feature pipelines, embedding pipelines, evaluation datasets and feedback loops.
- AI engineering leads and heads of AI, through retained search.
The LLM stack, specifically
A large share of current demand is retrieval-augmented generation and agentic workflows, and the difference between a demo and a production system is the engineering around the model. We recruit people who have built and operated these systems: RAG architectures over vector databases (Pinecone, Weaviate, Milvus, pgvector, FAISS), chunking and embedding strategy, reranking, grounding and citation, evaluation frameworks that catch hallucination and regression, guardrails, and orchestration with LangChain, LlamaIndex or custom frameworks. On the model side: supervised fine-tuning and parameter-efficient methods (LoRA, QLoRA), preference tuning (RLHF, DPO), distillation, and when fine-tuning is the wrong answer entirely.
What we screen for
Production, not notebooks. We build and run our own AI recruiting systems on our own hardware, in sync and in parallel - so the bar we screen to comes from operating these systems, not reading about them. We ask what shipped, what it serves, and what broke. Candidates should be able to talk about evaluation discipline - how they know the model got better and for whom - and about the economics: tokens, latency, GPU-hours and what the system costs to run. For research-track roles we look at publication and reproduction history; for product roles we look at owned outcomes in live systems. If someone is a prompt-curious engineer rebranding as an AI specialist, you will hear that from us before the interview.
Who we work with
Foundation-model and applied-AI startups, financial services and healthcare technology companies building AI into regulated products, enterprises standing up internal LLM platforms, and established software companies adding ML capabilities. Tooling changes every quarter; the screening questions do not.
New York and North Carolina
We recruit across the New York metropolitan area and from our base in Raleigh-Durham, where Research Triangle Park and the universities produce a steady flow of ML engineers and researchers, and we support remote and hybrid AI teams nationally.
Ways to hire
Contract, contract-to-hire or direct placement. We agree on the technical bar, the interview process and realistic timelines before the search starts. Current openings are on our jobs page.
Talk to us
Email patrick@triangleworkforce.com or call (984) 983-3100.
Fintech and financial services staffing · Healthcare IT staffing · Retained search · All recruiting services · Current open roles