AI Systems Builder (Impact-Driven) [33164]

Stealth Startup; Menlo Park (CA); On-site; USD 140,000 - 190,000; Full time

Job summary

Stealth Startup in Menlo Park, CA is hiring an Applied AI Engineer to design, fine-tune, and deploy production-grade AI models powering next-generation AI agents in financial workflows. The role sits at the intersection of research and real-world deployment — taking models from experimentation to high-stakes production systems, with collaboration across product and engineering teams globally.

Responsibilities include fine-tuning LLMs and multimodal models, building AI agents for multi-step

Qualifications

• Strong Python and ML framework experience for model development.
• Hands-on experience with PyTorch and multimodal models.
• Familiarity with evaluation and ablation studies in ML workflows.

Responsibilities

• Fine-tune LLMs and multimodal models for domain-specific use cases.
• Design post-training workflows (instruction tuning, RLHF, evaluation loops).
• Run ablation studies and hyperparameter tuning.
• Build AI agents capable of executing multi-step workflows.
• Integrate models into production environments.
• Improve robustness, accuracy, and reliability.

Skills

Python

PyTorch

Hugging Face

Production deployment

Experimentation & evaluation

Tools

Hugging Face ecosystem

GPU optimization

Job description

We’re hiring an Applied AI Engineer to design, fine-tune, and deploy production-grade AI models powering next-generation AI agents in financial workflows.

This role sits at the intersection of research and real-world deployment — taking models from experimentation to high-stakes production systems.

What You’ll Do

• Fine-tune LLMs and multimodal models for domain-specific use cases
• Design post-training workflows (instruction tuning, RLHF, evaluation loops)
• Run ablation studies and hyperparameter tuning

AI Agent Systems

• Build AI agents capable of executing multi-step workflows
• Integrate models into production environments
• Improve robustness, accuracy, and reliability
• Develop evaluation frameworks for edge cases
• Work with proprietary datasets
• Build labeling and data-quality pipelines
• Partner with product and engineering teams globally

Requirements

• 1–4 years hands-on ML/AI training experience
• Strong Python + PyTorch
• Experience with Hugging Face ecosystem
• Experience deploying models to production
• Understanding of model evaluation trade-offs

Bonus: RLHF, model distillation, RAG, distributed training, GPU optimization, multi-agent systems.