About the position

The ML Engineer on the Alerts Team plays a pivotal role in the intelligence layer that powers Samdesk’s automated alert pipeline - converting raw unstructured text to actionable crisis intelligence. You will own the quality of the output of our data pipeline. What does this mean? You will own the design and implementation of AI agents, orchestrate the interplay between our LLMs and data pipeline, and build the internal tooling that our operations, ML, and product teams rely on daily. You will work at the intersection of large-scale data systems and cutting-edge AI infrastructure, and your decisions will have a direct impact on system reliability and the quality of alerts delivered to users around the world. This role reports into the Alerts Team and collaborates closely with features, infrastructure, and product teams.

Responsibilities

• Design, build, and deploy ML models across the full lifecycle, from designing the ML architecture through error analysis and deployment
• Fine-tune and adapt LLMs using domain-specific alert data, including dataset curation, supervised fine-tuning, preference optimization, evaluation, and safe production rollout
• Upgrade models to newer versions, ensuring each new version measurably outperforms the last
• Work hands-on with Python ML libraries such as PyTorch, TensorFlow, Hugging Face, and XGBoost
• Collaborate with data and engineering teams to build scalable ML pipelines
• Contribute to data labeling strategies, feature engineering, and model evaluation frameworks
• Design and implement AI agents that coordinate LLM inference with our real-time data pipeline
• Build and maintain the orchestration layer governing how language models interact with structured pipeline outputs
• Integrate with OpenAI and Anthropic APIs, including prompt engineering, tool use, and response handling at scale
• Ensure agent workflows are observable, testable, and fault-tolerant in production
• Monitor and report on model performance, drift, and inference latency in production
• Attention to detail and problem-solving aptitude
• Set the bar for code and model quality through rigorous review of code, experiments, and evaluation results, and through mentorship
• Champion reproducibility through experiment tracking, versioned datasets, and robust evaluation so models and systems can be safely iterated on
• Decompose complex requirements into accurate effort estimates

Requirements

• 2+ years of professional experience in a machine learning or applied ML engineering role
• Familiarity with NLP, text classification, or information retrieval (a strong asset given our domain)
• Comfort working with large, noisy, real-world datasets
• Demonstrated experience building and operating AI agents or LLM-powered systems in production
• Hands-on experience with OpenAI and/or Anthropic APIs, including tool use, streaming, and prompt management
• Experience evaluating the outputs of ML components (ie precision and recall)

Nice-to-haves

• Experience with real-time data pipelines or event-driven architectures
• Familiarity with LLM evaluation frameworks, observability tooling, or RAG architectures
• Background in news, media monitoring, or open-source intelligence (OSINT)
• Solid working knowledge of AWS services (S3, SQS, CloudWatch) and ML infrastructure such as GPU-based inference, or vector databases

Benefits

• We encourage qualified applicants of all backgrounds including ethnicity, religion, disability status, gender identity, sexual orientation, family status, age, nationality, and education levels to apply.
• If you are contacted for an interview and require accommodation during the interviewing process, please let us know.