Multimodal AI Engineer, Document Understanding

Responsibilities

• Develop, train, and optimize machine learning models for document structure understanding, table extraction, layout analysis, and multimodal content processing
• Build robust data pipelines, evaluation frameworks, and experimentation infrastructure
• Design and implement production ML systems that handle complex, real-world documents at scale
• Stay current with latest advances in vision-language models, document AI, and multimodal learning
• Collaborate with engineering teams to integrate ML innovations into production APIs
• Contribute to both our open-source frameworks and enterprise offerings
• Drive technical decisions while balancing research exploration with product delivery

Required Qualifications

• 3-7 years of experience in machine learning engineering or applied research
• Strong software engineering fundamentals with production Python experience (modern tooling: uv, ruff, mypy, Pydantic)
• Hands-on experience training, fine-tuning, or deploying ML models in production
• Deep understanding of modern ML techniques, particularly in computer vision, NLP, or multimodal learning
• Experience with at least one of: data pipeline development, model training/fine-tuning, or ML infrastructure
• Ability to read and implement from research papers and technical specifications
• Track record of executing with high intensity in fast-paced environments
• Strong technical communication skills and comfort with open-source collaboration

Preferred Qualifications

• Experience with vision-language models, transformer architectures, or model fine-tuning (LoRA, QLoRA)
• Experience building evaluation frameworks, benchmarks, or data quality pipelines
• Experience with model serving frameworks (vLLM, TensorRT, ONNX) or MLOps tools
• Experience specifically with document understanding, OCR, or layout analysis
• Contributions to open-source ML projects or frameworks
• Experience with LLM applications and RAG systems
• Strong understanding of model optimization techniques (quantization, distillation, pruning)
• Experience with Docker/Kubernetes and distributed systems
• Active participation in ML research community