Staff Engineer, Learned Dynamics for Autonomy

Quiet Capital; Boston (MA); On-site; USD 185,000 - 325,000; Full time

Benefits offered by this job

Catered lunches

Snacks provided

Beverages on site

Job summary

Merlin Labs in Boston, MA, seeks a senior ML/AI researcher to design and train world models that predict aircraft state, environment, and other traffic. You will develop rollout tools for planners to evaluate candidate plans against predicted outcomes and drive uncertainty quantification for safety monitoring.

You will work with simulation and flight data to improve fidelity, compare against physics-based models, and ensure results are fully traceable to data and configurations.

Qualifications

• 10+ years building and training deep learning models.
• Experience with sequence, dynamics, video or trajectory prediction.
• Rigorous evaluation and uncertainty quantification.
• Familiarity with reading flight dynamics models and control concepts.

Responsibilities

• Design, train and evaluate world models predicting aircraft state and environment.
• Build rollout and imagination machinery for planner evaluation of candidate plans.
• Own calibrated predictive uncertainty and out-of-distribution signals for safety monitoring.
• Characterize validity boundaries across flight regimes, weather, density.
• Collaborate with Data/Sim to train in simulation and validate against flight data.
• Benchmark learned dynamics against physics-based models and report where classical methods win.
• Ensure reproducibility: trace results to data version, config, and seed.

Skills

Deep learning

Sequence modeling

Uncertainty quantification

Dynamical systems

Education

Bachelor's degree or higher in CS/AI/Data Science/Engineering

Tools

PyTorch

Job description

Merlin Labs in Boston, MA, seeks a senior ML/AI researcher to design and train world models that predict aircraft state, environment, and other traffic. You will develop rollout tools for planners to evaluate candidate plans against predicted outcomes and drive uncertainty quantification for safety monitoring.

You will work with simulation and flight data to improve fidelity, compare against physics-based models, and ensure results are fully traceable to data and configurations.