Forward Deployed Machine Learning Engineer

Role Overview

As a Forward Deployed Machine Learning Engineer at Federato, you will work directly on building, deploying, and iterating on machine learning models and agentic workflow features for real customer needs.

Responsibilities

• Build and deploy ML models and agentic workflow features for submission intake, underwriting decision-making, and automation tasks.
• Apply ML techniques to improve accuracy and overall system performance; ensure solutions are robust, reliable, and production-ready.
• Deploy and adapt autonomous agent behaviors into customer-specific workflows, translating core AI capabilities into practical solutions.
• Develop and maintain evaluation pipelines, monitoring systems, and performance metrics to ensure reliability as production conditions evolve.
• Monitor production using logs, metrics, and user feedback to diagnose issues, debug failures, and drive resolution.
• Take end-to-end ownership of problems by implementing fixes or coordinating with engineering and infrastructure teams.
• Partner with Data Science and Engineering to iterate quickly and deliver high-impact outcomes.

Requirements

• BS or MS in Mathematics, Operations Research, Data Science, Artificial Intelligence, or related field, with foundational knowledge in machine learning, deep learning, and natural language processing.
• 2+ years experience as a Machine Learning Engineer / Applied Scientist or similar, delivering ML solutions in production.
• Ability to translate business needs into technical solutions, including experience working with customers or stakeholders.
• Hands-on experience with adapting/extending/deploying ML/LLM systems, including agentic workflows and prompt engineering.
• Strong experimentation background with evaluation and monitoring pipelines, including analyzing production logs and debugging.
• Experience deploying and iterating on ML systems in cloud environments in collaboration with engineering teams.
• Proven track record of ownership and driving issues to resolution in production systems.

Nice to Haves

• The posting strongly emphasizes customer-facing deployment and reliability/monitoring ownership; additional depth in LLM evaluation/monitoring would likely be beneficial.