Hybrid AI Engineer

1. Role Title:

Hybrid AI Engineer (LLM, Vision Language Models)

2. Role Definition:

Hybrid AI Engineer responsible for building end-to-end AI systems across data, training, evaluation, and deployment, primarily for LLM use cases with targeted involvement in multimodal (VLM) workflows.

3. Key Responsibilities:

1. Data Engineering & Curation:

- Build and manage data pipelines, including dataset preparation, filtering, deduplication, and preprocessing for ML workflows.

2. Model Training & Fine-Tuning:

- Train and fine-tune models using techniques such as LoRA, SFT, and DPO for LLM-based systems.

3. AI Systems Development:

- Develop RAG pipelines, evaluation frameworks, and benchmarking systems for applied AI use cases.

4. Deployment & Optimization:

- Deploy models using SageMaker, EKS, vLLM, or NIM and optimize for latency, throughput, and cost.

5. Multimodal Workflow Support:

- Work with document and image-based datasets, supporting OCR, document AI, and multimodal training/inference workflows.

4. Must Have Skills:

1. ML Engineering & Model Training:

- Strong Python skills with hands-on experience in PyTorch/Hugging Face and LLM fine-tuning techniques (LoRA, SFT, DPO).

2. Data Pipelines & Dataset Engineering:

- Experience building and managing data pipelines, including dataset preparation, filtering, and preprocessing.

3. LLM Systems & Evaluation:

- Practical understanding of RAG systems, model evaluation, benchmarking, and applied AI workflows.

4. Applied AI Problem Solving:

- Ability to design, debug, and optimize end-to-end ML systems across data, training, and inference stages.

5. Multimodal Data & VLM Awareness:

- Familiarity with image-text/document datasets, OCR/document AI workflows, and basic multimodal concepts.

5. Years of Experience:

- 25 years of ML / data engineering experience.

- Hands-on experience (624 months) with LLMs preferred.