SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Armada is a hyperscaler for edge AI infrastructure, backed by ~$500M in funding from Founders Fund, Lux, BlackRock, and Microsoft, with strategic partnerships including Microsoft, Dell, Palantir, NVIDIA, SpaceX, and Skydio. The company delivers modular AI infrastructure for sovereign and edge computing, deployed across 60+ countries in energy, defense, and critical infrastructure.
As an AI Engineer, you will own substantial portions of the AI lifecycle for production systems deployed in extreme environments: offshore energy platforms, remote mines, defense installations, transportation networks, and autonomous systems. Your work spans multimodal and generative AI, real-time computer vision, large language and vision-language models, robotics, reinforcement learning, time-series analysis, and anomaly detection.
Key responsibilities include translating operational problems into AI requirements and production architectures; designing, training, fine-tuning, and deploying models for vision, language, and multimodal tasks; building training and evaluation datasets from video, images, text, telemetry, and sensor data; optimizing inference through quantization, pruning, distillation, and hardware-aware acceleration; building reliable AI services with Python, testing, versioning, and observability; deploying containerized AI workloads across Kubernetes, cloud, on-premises, and disconnected edge environments; building monitoring and retraining pipelines; and partnering with customers, product teams, and domain experts to move prototypes into production.
You will diagnose issues across data pipelines, models, GPUs, orchestration, and networking. The role requires at least three years of relevant industry experience beyond experimentation and prototyping, combining rigorous ML/AI fundamentals with strong software-engineering judgment. You should be equally comfortable interpreting research, implementing new methods, diagnosing performance bottlenecks, and deploying resilient services across heterogeneous environments with real-world constraints: intermittent connectivity, strict latency requirements, limited compute, noisy data, and rigorous security demands.
