Engineering Manager

Shield AI is a venture-backed defense-tech company with the mission of protecting service members and civilians with intelligent systems. Its products include Hivemind autonomy software, V-BAT and X-BAT aircraft, and Aechelon simulation and synthetic reality technologies. With offices and facilities across the U.S., Europe, the Middle East, and Asia-Pacific, Shield AI’s technology actively supports operations worldwide. For more information, visit www.shield.ai. Follow Shield AI on LinkedIn, X, Instagram, and YouTube. 

Job Description:

Lead the future of physical AI in defense by building the reasoning models for unmanned autonomous systems, across aerial, maritime, land, and space domains. 

We are looking for a leader with not only the deep technical expertise or the abilities to build a world-class team, but also someone with the strong conviction that AI will protect service members and civilians. This role is responsible for building and leading a team that researches and designs models, establishes training pipelines and rigorous evaluations, and ensures mission effectiveness at the edge under real-world constraints. These models must support individual vehicles and coordinated teams of hundreds of agents, while meeting the constraints of operating at the edge in contested environments. 

In this role, you’ll partner closely with world-class teams in perception, MLOps, large-scale simulation, and inference to deliver capabilities as part of the broader Hivemind Enterprise team. 

What success looks like: your team’s reasoning models deliver exceptional outcomes for baseline platforms and missions, which can then be rapidly and confidently adapted to new platforms and missions by our solutions engineering organisation. 

What you'll do:

• Build & lead the team that designs, trains, adapts, and evaluates reasoning and cognition models intended for deployment across our unmanned platforms. 
• Define the technical roadmap for single-agent and multi-agent models, including architecture choices, training methods, data strategy, and evaluation criteria. 
• Make informed tradeoffs among foundation-model approaches (e.g. VLMs, VLAs), specialized deep-learning models, and traditional AI/ML methods. 
• Advance model capabilities for interpreting context, reasoning over time, coordinating among agents, and adapting to incomplete or changing information.