Sovereign AI Model Engineer – Geospatial Autonomy

NestAI; Tampere; On-site; EUR 80,000 - 120,000; Full time

Benefits offered by this job

Occupational healthcare (Mehiläinen)

Epassi Flex benefit

Lunch benefit

Regular team lunches and all-hands

Mac or PC of your choice

Holiday allowance (50% of holiday pay)

Phone benefit per company policy

Job summary

NestAI invites a senior Model Engineer to join our initiative building sovereign AI models for battlefield autonomy and orchestration. You will work on models that adapt to changing operations and leverage simulated and real data in a modular platform.

You'll prototype and benchmark multimodal and geospatial reasoning solutions, create early evaluation harnesses, and contribute technical judgment to steer the project. The preferred location is Finland, with flexibility for other EU countries.

Qualifications

• Solid hands-on experience training, fine-tuning, and evaluating ML models - ideally including multimodal or large-scale model work.
• Comfort working in a still-forming, under-specified environment: this project doesn't have settled tooling, data, or team structure yet, and the ability to make progress without waiting for all of that to exist is more valuable than following an established playbook.
• Genuine curiosity about geospatial/spatial reasoning, simulation, or embodied AI problems - these show up constantly in this domain and current foundation models handle them poorly.
• Nice to have: a PhD in a relevant quantitative field; prior exposure to autonomous systems, simulation, or edge-constrained systems; experience or interest in defence, dual-use, or other regulated/sovereign-technology environments.

Responsibilities

• Build and iterate on model prototypes across both tracks: from early proof-of-life experiments to benchmarking against existing C2/Edge functionality.
• Work hands-on with multimodal and geospatial reasoning problems: LLMs currently fall short on spatial/geospatial reasoning, and closing that gap is core to the effort (e.g. representing maps and terrain as something a model can actually reason over, not just look at).
• Help stand up early evaluation harnesses and benchmarks so progress is measurable against a real baseline, not just a demo.
• Work against simulated/synthetic data in the near term, coordinating with whoever owns the data owners on what's usable and what needs to be built.
• Contribute technical judgment upward: flag when a direction isn't working rather than quietly grinding on it, making sure we fail fast.

Job description

NestAI invites a senior Model Engineer to join our initiative building sovereign AI models for battlefield autonomy and orchestration. You will work on models that adapt to changing operations and leverage simulated and real data in a modular platform.

You'll prototype and benchmark multimodal and geospatial reasoning solutions, create early evaluation harnesses, and contribute technical judgment to steer the project. The preferred location is Finland, with flexibility for other EU countries.