Role Description & Responsibilities
Machine Learning
About Scale AI
Scale AI is a high-growth technology leader building scalable solutions for global users. Millions of companies and teams rely on their infrastructure and services to execute high-value workflows, analyze massive datasets, and scale modern operations.
CadetX Candidate Advantage: As an applicant for this Machine Learning role, you are evaluated on practical proficiency with real-world query execution, data pipelines, and clear stakeholder communication.
CadetX Role Brief & Company Specifications
Role Mission & Impact
The goal of a Machine Learning Engineer at Scale is to leverage techniques in the fields of generative AI, computer vision, reinforcement learning, and agentic AI to improve Scale's products and customer experience in production environments. Our machine learning engineers take advantage of robust internal infrastructure and unique access to massive datasets to deliver improvements to our customers.
Our Public Sector Machine Learning team is focused on deploying cutting-edge models to mission-critical government systems through products like Donovan and Thunderforge . Our work spans multiple modalities, with a strong focus on both large language models and computer vision. On the LLM side, we are developing agentic systems that help solve complex operational and planning challenges for government partners. This includes building agent frameworks that integrate with custom retrieval pipelines and production APIs, as well as evaluation tools to benchmark and refine agent behavior. We're also advancing research in areas like reinforcement learning for agentic LLMs, with successful deployment into real-world operational environments. On the computer vision front, we're training advanced models to increase labeling throughput and automate perception tasks. Our efforts include building large-scale fine-tuning pipelines, training models across multiple modalities, and developing generalizable vision foundation models to support a wide range of defense applications.
Key Responsibilities & Day-to-Day Scope
- ✓ Take state of the art models developed internally and from the community, use them in production to solve problems for our customers and taskers
- ✓ Improve and maintain production models through retraining, hyperparameter tuning, and architectural updates, while preserving core performance characteristics
- ✓ Collaborate with product and research teams to identify and prototype ML-driven product enhancements, including for upcoming product lines
- ✓ Work with massive datasets to develop both generic models as well as fine tune models for specific products
- ✓ Build scalable machine learning infrastructure to automate and optimize our ML services
- ✓ Serve as a cross-functional representative and advocate for machine learning techniques across engineering and product organizations
- ✓ Be comfortable learning new technologies quickly and managing multiple priorities in a fast-paced environment
- ✓ Comfortable with light travel (approximately 10%) for customer interaction and team needs
- ✓ 2+ years of experience building and deploying applied ML systems in production environments
Official Qualifications & Technical Requirements
- • Extensive experience with GenAI, Agentic AI, natural language processing, deep learning and deep reinforcement learning, or computer vision in a production environment
- • Solid background in algorithms, data structures, and object-oriented programming
- • Strong programming skills in Python, experience in Tensorflow or PyTorch
- • Experience deploying software into environments you can't reach from your laptop - on-prem, edge, air-gapped, or otherwise restricted networks. Regulated industries count; the constraint is the point, not the sector.
- • Any prior exposure to government or defense work: military or civilian service, a cleared internship, or time at a federal contractor.
- • Hands-on fine-tuning of open-weight models - LoRA/PEFT, instruction tuning, or training embedding models, at work or on your own.
- • Having written evaluations for a system whose output isn't deterministic: benchmarks, LLM judges, or a regression suite that caught something real.
- • Experience with geospatial data or maps - GIS tooling, spatial reference systems, or imagery.
- • A shipped project with real users behind it, where you owned it after launch rather than handing it off at merge.
Preferred Skills & Technology Stack
- + Familiarity with MLOps pipelines (Docker, Kubernetes, Kubeflow, Triton Inference Server).
- + Knowledge of distributed training techniques and GPU acceleration (CUDA).
- + Relevant coursework, research publication, or open-source contribution in ML/DL.
Tech Stack & Tools in this role
Expected technical competencies for Day 1 productivity:
★ Life & Culture at Scale AI
Verified EmployerScale AI champions rigorous analytical discipline, continuous learning, and low-ego teamwork. Team members enjoy flexible hybrid allowances, learning stipends, and structured mentorship to fast-track career mobility.