Senior Forward Deployed Engineer, Open Models

This is a fully remote position, open to applicants in California.

📋 Description

• Lead the development of AI solutions from initial discovery and scoping through to post-training activities, harness optimization, and deployment.

• Formulate post-training and harness strategies utilizing appropriate methodologies, data, assessments, and success metrics.

• Create evaluations and construct synthetic data pipelines.

• Conduct post-training using LoRA and other parameter-efficient techniques, supervised fine-tuning, as well as both synchronous and asynchronous reinforcement learning.

• Make informed decisions regarding architecture and hyperparameters to enhance outcomes.

• Optimize prompting, tool utilization, orchestration, and context management for superior domain-specific performance.

• Enhance inference processes for economical production operations.

• Spearhead technical engagements by defining project scopes, crafting statements of work, and overseeing delivery alongside partner engineering teams and joint customers.

• Develop reusable recipes, reference pipelines, and playbooks for partners.

• Link NVIDIA model research to practical applications for partners and customers.

• Collaborate with product, research, and engineering teams to integrate field insights into Nemotron, Cosmos, and NVIDIA's software ecosystem.

⛳️ Requirements

• Bachelor's, Master's, or PhD in Computer Science, Engineering, Machine Learning, or a related discipline, or equivalent experience.

• Over 8 years of experience in ML or software engineering, applied AI, FDE, or solutions architecture.

• At least 3 years of practical experience with LLM post-training and deploying models into production.

• Strong foundational knowledge in deep learning and machine learning principles.

• Extensive expertise in LoRA/PEFT, SFT, and reinforcement learning, including both synchronous and asynchronous methods.

• In-depth understanding of model architectures and hyperparameter tuning.

• Awareness of large-scale distributed training and strategies for optimizing inference performance.

• Capability to configure and optimize workloads for memory efficiency, throughput, latency, and scalability across multi-GPU and multi-node environments.

• Experience with synthetic data pipelines, dataset curation and filtering, and evaluation sets.

• Background in harness engineering for LLM and agentic systems.

• Proficiency in PyTorch, Hugging Face Transformers, Python, and Linux.

• Demonstrated ability to define project scopes, shape statements of work, and lead delivery efforts with partners and clients.

• Exceptional organizational skills and the ability to facilitate complex decision-making across teams and organizations.

• Proficient in conveying intricate technical concepts to engineers as well as C-level executives.

• Strong analytical abilities, a sense of intellectual curiosity, and a dedication to the success of partners and customers.

🏝️ Benefits

• Equity

• Benefits