Sovereign AI Model Development Lead

<h3 id="docs-internal-guid-4bc64443-7fff-59d9-5ee8-b308a853b61a"><span>About NestAI</span></h3><p><span>NestAI is a European AI lab for defence delivering the adaptive operating system for modern battlefield operations. We develop adaptive intelligence for unmanned and command systems: AI that continuously learns from operational data and adapts to changing conditions.</span></p><p><span>NestOS is our open, modular and interoperable platform that enables this at scale and keeps capability evolution under sovereign European control. Founded in 2025 by Peter Sarlin, whose previous company Silo AI became Europe's largest AI acquisition, NestAI brings together over 200 engineers and scientists and partners with leading European defence forces and industry.</span></p><h3><span>Role Description</span></h3><p><span>Europe builds the unmanned vehicles and the platforms. It does not build the models that fly and command them. The intelligence inside Europe’s defences runs on models trained elsewhere, and access to the best models can be cut off overnight. NestAI is </span><a target="_blank" href="https://www.nestai.com/blog-posts/nestai-builds-sovereign-ai-for-european-defence"><span>building sovereign AI models for defence to change this</span></a><span>. We are initially focusing on models for battlefield autonomy and orchestration, bringing together real operational data, simulation and synthetic data to build models that can continuously adapt as the battlefield changes.</span></p><p><span>We're hiring a Model Development Lead to help us drive the effort. This is a senior technology leadership role where you'll set technical and architectural direction for an existing team, making the calls on what to build and growing the team around you.</span></p><p><span>You will chart what a large-scale battlefield foundation model could become, looking beyond the requirements of today’s product roadmaps. This includes what battlefield data it should learn from, what reasoning capabilities it should develop, how it could support strategic, operational, and tactical decision-making and how it could adapt to new tactics emerging from the field.</span></p><p><span>In order to succeed, you need deep technical competence in model development, including pre-training, fine-tuning, evaluation, multimodal learning, long-context reasoning, model compression, and understanding of deployment trade-offs.</span></p><h4><span>What you'll do</span></h4><ul><li><p><span>Own the technical architecture and roadmap, translating an ambitious vision into a sequenced plan that can be executed and that creates verifiable milestones.</span></p></li><li><p><span>Build an understanding of NestAI strategy of battlefield autonomy and orchestration and map how the model development acts as a layer that elevates the overall company roadmap through the foundation model.</span></p></li><li><p><span>Operate with real decision authority to make sure that we iterate quickly and fail fast. You’re not just an advisory voice but charted with making ambiguity reality.</span></p></li><li><p><span>Represent the project's technical thinking to help attract the caliber of talent this kind of work requires.</span></p></li><li><p><span>Report progress in a form leadership can act on: an agreed scoping, a phased plan with clear decision gates, and an honest account of risks and opportunities.</span></p></li><li><p><span>Grow and staff the team around so that you and the company are set out for success.</span></p></li></ul><h4><span>What we're looking for</span></h4><ul><li><p><span>Senior experience in training and leading the training of large-scale, multimodal AI models- ideally from a frontier AI lab (e.g. Anthropic, OpenAI, Google DeepMind) or a comparably serious scaled effort.</span></p></li><li><p><span>You have deep technical competence in model development, including pre-training, fine-tuning, evaluation, multimodal learning, long-context reasoning, model compression, and deployment trade-offs.</span></p></li><li><p><span>Experience building or adapting models for constrained deployment environments, including quantisation, distillation, pruning, or other compression techniques for edge, embedded, or air-gapped systems.</span></p></li><li><p><span>Strong evaluation discipline: ability