Principal AI/ML Architect

The Principal Data Scientist (ML/DL) drives the design and development of advanced AI solutions for medical applications. This role collaborates with senior data scientists, ML engineers, software engineers, and subject matter experts to enhance current technologies and lead forward-looking AI innovations.

Key responsibilities include improving existing NLP systems and extending them into cloud-based applications, with a strong focus on novel machine/deep learning techniques for information extraction and synthesis. The role spans the full lifecycle from PoC development to performance evaluation and production deployment leveraging statistical methods, deep learning, and LLM technologies.

The ideal candidate demonstrates a strong publication record in ML/DL (especially NLP, IR, and IE), deep familiarity with modern research, and proven experience deploying scalable ML solutions in production environments.

Key Responsibilities

• Lead end-to-end training and fine-tuning of LLMs (open-source: Qwen, LLaMA, Mistral; closed-source: Open AI, Gemini, Anthropic)
• Apply deep expertise in ML/DL frameworks (transformers, state space models, LLMs, agentic systems)
• Develop algorithms for text processing, information extraction, and retrieval
• Architect Graph RAG pipelines with knowledge graph–based contextual retrieval
• Design and optimize semantic/dense embeddings for search and document understanding
• Build advanced semantic retrieval systems (segmentation, indexing strategies)
• Scale distributed training (NCCL, InfiniBand, multi-GPU/multi-node)
• Apply RL techniques (RLHF, RLAIF) for alignment with human/domain goals
• Develop hybrid NLP systems (symbolic + ML approaches)
• Translate business needs into scalable AI solutions and production deployments

Preferred Qualifications

• PhD/Master’s in Computer Science, ML, or related field
• 11–15+ years in applied AI/ML with strong production track record
• Deep expertise in NLP, ML, DL, transformers, state-space architectures
• Strong Python, SQL, data prep, EDA skills; experience with PyTorch
• Cloud experience (Azure ML, AWS)

Technical Expertise:

• LLM training & fine-tuning (GPT, LLaMA, Mistral, Qwen)
• Graph RAG, knowledge graphs
• Embeddings (BGE, E5, SimCSE)
• Vector DBs (FAISS, Weaviate, Milvus)
• Document processing (OCR, layout parsing)
• Distributed training (NCCL, Horovod, Deep Speed)
• Networking (InfiniBand, RDMA)
• Model ensembles (stacking, boosting, gating)
• Optimization (Bayesian, PSO, Genetic Algorithms)
• Symbolic AI, rule-based systems
• Meta-learning, Mixture of Experts
• RL methods (RLHF, PPO, DPO, GRPO), SFT, LoRA, QLoRA, Axolotl
• Prompt optimization (Auto Prompt, Greater Prompt, DSPy), GEPA