3 months ago
Remote, United StatesSenior / Staff+
Base Salary
$200k - $400k/yr
Responsibilities
- Design, train, and evaluate multimodal foundation models for multi-robot mission execution.
- Drive model architecture, data mixture, curriculum, and training recipe decisions through rigorous experimentation.
- Curate and generate large-scale training datasets from synthetic and real-world data engines.
- Develop memory, communication, planning, and tool-use capabilities for AI agents operating across robot teams.
- Design benchmarks, metrics, and testing methodologies for rapid model iteration.
- Build and improve simulation environments for synthetic data generation, evaluation, and reinforcement learning.
- Translate foundational research into deployable, real-time perception and decision-making systems.
- Integrate AI systems with onboard autonomy stacks across engineering, robotics, and mission teams.
- Conduct field trials and mission operations to validate performance under real-world constraints.
Requirements
- At least 6 years of hands-on experience building and deploying AI models, including at least 2 years with multimodal models or agentic AI systems.
- Proficiency in Python, PyTorch, and modern machine learning infrastructure.
- Experience training, fine-tuning, and evaluating large-scale deep learning models with distributed compute infrastructure.
- Demonstrated ability to improve model performance through data curation, training strategy, evaluation design, and rigorous experimentation.
- Knowledge of SFT, DPO, and RLVR.
- Ability to take research from prototype to product in fast-moving environments.
- Bachelor’s, master’s, or doctoral degree in computer science, engineering, mathematics, physics, or a related technical field, or equivalent practical experience.
- U.S. person status required for access to export-controlled information or facilities.
- Advanced graduate research in AI, robotics, or machine learning is a plus.
- Experience building closed-loop simulation, evaluation, or reinforcement learning environments for agentic or physical AI systems is a bonus.
Benefits
- Meaningful equity and competitive PTO with a company holiday calendar.
- Premium medical, dental, and vision plans with $0 paycheck contribution.
- Unlimited AI tokens, catered lunch daily, and a fully stocked kitchen.
- EV charging and relocation assistance depending on role eligibility.
Categories
About Scout AI
Scout AI builds defense-focused AI software for unmanned ground, air, sea, and space systems. Its core product, the Fury foundation model, enables operators to command robotic platforms in natural language and coordinate autonomous behaviors; Scout also integrates the software with fielded payloads and compute. Founded in 2024 and headquartered in Sunnyvale, California, the privately held company is venture-backed by investors including Booz Allen Ventures and Draper Associates.