SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 235,200 - 294,000 / annual
Scale is seeking a Senior Machine Learning Engineer to own the design and deployment of generative AI, agentic AI, computer vision, and reinforcement learning systems for mission-critical government applications. You will work on the Public Sector ML team, focusing on products like Donovan and Thunderforge that bring cutting-edge models to government systems.
In this role, you will have design authority over a capability area within your team, setting technical patterns and ensuring they work under real-world constraints including classified environments, limited compute, and strict correctness requirements. You will own agent capabilities end-to-end—from architecture and implementation through evaluation—and define new patterns in problem spaces without established approaches. You'll take state-of-the-art models and deploy them to production, improve and maintain them through retraining and architectural updates, and build agent-level evaluation benchmarks and LLM judges to drive performance improvements.
You will partner with product and research teams to scope high-impact initiatives, build scalable ML infrastructure, and work directly with government users and subject-matter experts to translate their needs into technical direction. As a force multiplier for your team, you'll mentor at least one engineer, serve as a primary code reviewer, and act as your manager's go-to on feasibility questions. You'll communicate technical tradeoffs clearly to non-technical stakeholders and treat security and compliance as design constraints rather than blockers. The role involves approximately 10% travel for customer interaction and team needs, and requires an active security clearance.
The work spans multiple modalities with primary focus on agentic systems built on large language models. You'll develop agent frameworks integrating custom retrieval pipelines and production APIs, memory and context-management systems for long-running tasks, geospatial reasoning over maps and spatial data, and evaluation tooling to benchmark and refine agent behavior. You'll also apply reinforcement learning where appropriate and advance computer vision work for evaluation, labeling efficiency, and multimodal model training in defense applications.
REQUIREMENTS:
- 5+ years of experience building and deploying applied ML systems in production environments
- Extensive experience with GenAI, Agentic AI, natural language processing, deep learning, deep reinforcement learning, or computer vision in production
- Track record of owning architectural decisions and defending tradeoffs—not just implementing handed-down designs
- Experience shipping agentic systems with real production traffic and evaluation rigor (not prototypes or demos)
- Solid background in algorithms, data structures, and object-oriented programming
- Strong Python programming skills with experience in PyTorch or TensorFlow
- Experience mentoring or reviewing other engineers' work
NICE TO HAVES:
- Graduate degree in Computer Science, Machine Learning, or Artificial Intelligence
- Experience with cloud platforms (AWS, GCP) and deploying ML models in cloud environments
- Experience with computer vision, generative AI models, large language models, or agentic systems
- Familiarity with ML evaluation frameworks and agentic model design
- Experience deploying ML in classified, air-gapped, or IL5+ environments
- Geospatial or GEOINT experience
- Inference optimization experience
- Fine-tuning experience (SFT, RL, or embedding models)
