: Sr AI and Machine Learning Engineer :: Remote

Location: USA (Remote)

Duration: 6 months minimum

Exp: 12+ years

 

Skill: LLM, Machine Learning, PyTorch and/or Tensorflow

Role

We are looking for a senior-level engineer with deep expertise in large-scale ML/LLM systems. You will design, optimize, and deploy high-performance training and inference pipelines; build retrieval-augmented generation (RAG) architectures; and own end-to-end ML workflows across research, experimentation, and production. This role requires strong systems thinking, hands-on capability with modern ML frameworks, and the ability to independently drive architectural decisions.

 

Responsibilities

• LLM Training & Fine-Tuning
  Execute supervised fine-tuning, preference optimization, and evaluation workflows.
  Build reproducible training pipelines using PyTorch/TensorFlow, DeepSpeed, FSDP, or distributed training stacks.
  Implement evaluation harnesses for task-specific and generative metrics.
• Retrieval & RAG System Architecture
  Design embeddings strategies (dense, sparse/hybrid; instruction-tuned embeddings).
  Build vector search pipelines using FAISS, Milvus, or Pinecone with optimal indexing, sharding, and recall/latency tradeoffs.
  Develop production-grade RAG pipelines with prompt templates, context optimization, caching, and guardrails.
• Inference Optimization
  Implement low-latency inference flows using ONNX Runtime, TensorRT, vLLM, or other optimized runtimes.
  Apply quantization (INT8/FP8/4-bit), model distillation, graph optimizations, and batching strategies.
  Build and operate model-serving stacks (TorchServe, BentoML, custom GRPC microservices).
• Applied ML Systems Engineering
  Build ML workflows across structured/semi-structured/unstructured data.
  Develop feature pipelines, embeddings services, and data preprocessing stacks.
  Own CI/CD, experiment tracking, artifact management, and production deployment.
• Security, Compliance, and Architecture
  Implement secure data pipelines: PII redaction, encryption, access controls, lineage/auditability.
  Design privacy-preserving ML workflows (differential privacy, secure enclaves, RBAC/ABAC).
  Make architectural decisions around storage formats, compute orchestration, containerization, and observability.

 

Required Skills and Experience

• 6+ Years of relevant work experience
• Expert-level experience with LLMs, fine-tuning, evaluation, and scalable deployment.
• Strong proficiency in PyTorch or TensorFlow, plus Hugging Face Transformers/PEFT, LangChain, and vector DBs (FAISS, Milvus, Pinecone).
• Strong background designing retrieval systems, RAG architectures, embeddings, and hybrid search strategies.
• Hands-on experience with model acceleration frameworks: TensorRT, ONNX Runtime, vLLM, DeepSpeed, FSDP, etc.
• Ability to own and deliver complete ML systems—data pipelines → training → evaluation → serving → monitoring.
• Strong engineering fundamentals: Python, distributed systems, performance tuning, containerization, cloud compute.
• Ability to operate independently, make architectural calls, and drive complex ML/LLM projects without heavy guidance.

 

 

 

 

Navneet Singh US Technical Recruiter, Fast Hire INC  navneet@fasthireinc.com