Senior AI Autonomy Engineer — Edge & Vision-Language

AlleyCorp; San Jose (CA); On-site; USD 195,000 - 235,000; Full time

Job summary

Archer, an aerospace company based in San Jose, CA, seeks a Staff Engineer, AI Autonomy to advance intelligent, learned behavior for eVTOL systems. You will develop multi-model systems, fine-tune foundation models, and validate performance across simulation, hardware, and flight tests.

You will lead edge deployments, work with Safety and Simulation teams, and push models from prototype to deployed components under real latency and power limits.

Qualifications

• 8+ years of relevant experience.
• MS/PhD or equivalent in CS/Robotics; exceptional BS candidates considered.
• Hands-on experience with vision-language/VLA models.
• Strong grasp of transformer-based multimodal models and fine-tuning.
• Experience with model-serving integration end-to-end.
• Real 0→1 project experience and comfort with ambiguity.

Responsibilities

• Develop and integrate multi-model systems (VLN/VLA, perception, language).
• Design benchmarks, report comparisons to baselines and results.
• Fine-tune foundation models; weigh tradeoffs clearly.
• Validate with data-driven tests in simulation, hardware, and flight.
• Deploy and monitor models on edge hardware under real constraints.

Skills

Vision-language models

Multimodal transformer architectures

Model fine-tuning and integration

Edge deployment / latency constraints

Ambiguity handling / startup mindset

RL/Imitation/Self-supervised learning

Education

MS/PhD in CS/Robotics or related

Tools

PyTorch

Job description

Archer, an aerospace company based in San Jose, CA, seeks a Staff Engineer, AI Autonomy to advance intelligent, learned behavior for eVTOL systems. You will develop multi-model systems, fine-tune foundation models, and validate performance across simulation, hardware, and flight tests.

You will lead edge deployments, work with Safety and Simulation teams, and push models from prototype to deployed components under real latency and power limits.