SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Merlin Labs (NASDAQ: MRLN) is a publicly traded aerospace and defense company building autonomous flight systems. The company has proven its non-human pilot autonomy platform through hundreds of autonomous flights from facilities in New Zealand, Rhode Island, and Massachusetts. Merlin is expanding to accelerate development and deployment of its autonomy platform for commercial and defense aviation.
In this Staff Engineer role, you will own the learned predictive models of aircraft dynamics, environment, and other actors—and the rollout machinery that turns those models into evaluated candidate futures for planning.
Key Responsibilities:
- Design, train, and evaluate world models that predict aircraft state, environment, and other traffic evolution under candidate action sequences
- Build rollout and imagination machinery enabling planners to evaluate candidate plans against predicted outcomes before execution
- Own uncertainty quantification: deliver calibrated predictive uncertainty and out-of-distribution signals consumed by the safety monitoring layer
- Rigorously characterize model validity boundaries—flight regimes, weather, traffic densities, and configurations where predictions are trustworthy versus degraded
- Work the sim-to-real gap bidirectionally with the Data/Sim team: train in simulation, validate against flight data, feed discrepancies back into simulator fidelity
- Benchmark learned dynamics against Merlin's physics-based and flight-controls models; honestly assess where classical approaches outperform learned ones
- Contribute to shared evaluation harness and enforce reproducibility standards: every result traceable to data version, config, and seed
Requirements:
- Degree in Computer Science, Artificial Intelligence, Data Science, Computer Engineering, Applied Math, or related field
- 10+ years building and training deep learning models with meaningful work on sequence, dynamics, video, or trajectory prediction
- Strong PyTorch expertise and solid grasp of training stack (distributed training, mixed precision, experiment tracking, debugging)
- Working knowledge of dynamical systems, state estimation, or control; ability to read flight dynamics models and understand what your network replaces
- Demonstrated rigor in evaluation and uncertainty quantification
- Comfort with messy real-world sensor and telemetry data
Nice to Have:
- Model-based RL, latent dynamics models, or learned simulators
- Trajectory prediction experience in autonomous driving or robotics
- Aerospace background (flight dynamics, aircraft performance, air traffic)
- Multimodal fusion across vision, state, and structured mission context
