• Design and implement ML pipelines (training, validation, deployment, and inference) with a focus on scalability and cost. • Deploy models to production using MLOps practices (versioning, reproducibility, automation, and performance/drift monitoring). • Build and maintain inference services (batch and/or real-time), APIs, and integrations with internal systems. • Collaborate with data scientists, analysts, and product teams to translate business needs into solutions. • Ensure quality (testing, code reviews, documentation), security, and engineering best practices. • Track operational and outcome metrics (SLA, latency, cost, accuracy, business impact).

• Design and build scalable ML services for enrichment workflows, including model training pipelines and high-performance inference APIs • Deploy and optimize models using modern inference libraries and frameworks (ONNX, vLLM, TensorRT, etc.) to achieve low-latency, high-throughput performance • Collaborate with software engineers and product teams to define data requirements, feature engineering strategies, and model evaluation metrics • Build robust monitoring, observability, and evaluation systems to ensure model quality and service reliability in production • Stay current with emerging ML techniques, tools, and best practices, particularly in areas like model optimization, efficient inference, and large-scale data processing

• Create high-quality instructional materials that teach modern data annotation and labeling workflows used to train and evaluate machine learning systems. • Develop content across modalities—including text, image, video, audio/speech, and cross-modality labeling. • Collaborate with internal program team to produce session plans, hands-on practice, examples, and assessments aligned to defined learning outcomes. • Content should model best-in-class annotation behavior: rule-based decisions, clear rationales, consistent application of guidelines, and appropriate handling of ambiguity. • Produce complete content packages per session/module, such as facilitator guide, learner materials, demos or walkthroughs, practice activities, strong vs. weak examples, rubrics/evaluation criteria, and short knowledge checks/assessments. • Create hands-on exercises using a common annotation platform, including task setup guidance, labeling instructions, example labels, edge cases, and review workflows. • Review and iterate on content based on internal feedback and peer review. • Participate in weekly and ad hoc meetings with the program team to align on objectives, standards, and delivery constraints. • Support content deployment into our learning platform (training provided).

• Build and maintain evaluation frameworks for document models, LLMs, OCR, and structured extraction. • Define metrics, benchmarks, and validation strategies for real-world document workloads. • Design and curate high-quality datasets for supervised training, fine-tuning, and validation. • Create scalable preprocessing pipelines for PDFs, scans, images, forms, and semi-structured documents. • Train and fine-tune transformer-based OCR, VLMs, layout models, and open-source LLMs for document understanding tasks. • Optimize models for reliability, accuracy, and cost efficiency in production environments. • Deploy ML models with modern inference runtimes (vLLM, TGI, TensorRT, ONNX Runtime). • Build guardrails, monitoring, and fallback mechanisms to ensure safe and predictable model behavior. • Develop retrieval and chunking strategies tailored to document structures (tables, forms, multi-page PDFs). • Optimize end-to-end RAG pipelines for semantic search, Q&A, and workflow automation. • Partner with PMs, backend engineers, and product designers to define AI opportunities and translate requirements into technical solutions.