Senior Software Engineer - Machine Learning

You'll be a generalist responsible for building and running large-scale data, machine learning, and agentic systems. The focus is operational ML/AI, including agentic systems and geospatial data pipelines.

You should be comfortable owning the full lifecycle: from data ingestion and distributed processing to model development, deployment, and monitoring. This role requires the ability to iterate quickly from initial concept to a robust, production-ready solution.

Key Responsibilities

Take ownership of the end-to-end AI/ML lifecycle, with a strong focus on dealing with complex and messy data, thorough evaluation of different approaches, and successfully deploying robust models, and handling cost vs performance tradeoffs

Implement and integrate large-scale, agent-based systems with access to external systems, building these solutions from the ground up and integrating them with our existing infrastructure

Establish observability for pipelines, models, and agents (metrics, tracing, alerting)

Collaborate with product and customer teams to drive revenue

Requirements

Strong experience with distributed data processing, particularly Spark and SQL

Proven expertise in building production machine learning systems, including working with large, wide datasets, effective training, deployment, and monitoring

Experience designing and deploying task-oriented AI agents and working with coding agents

Experience working with cloud services across data, compute, and ML

Strong communication abilities, including code architecture and documentation, at a level where any technical team member can troubleshoot and contribute easily

Languages: Scala, Python

Tools / Frameworks: Spark, AWS Sagemaker / Bedrock, Kubernetes

Nice to Haves

Startup experience or growing projects from 0 to production in a larger org

Experience with large geospatial datasets, formats, and indexing strategies

Experience building operational AI agents that work at scale (millions of separate, complex tasks including web research)

Experience with fine-tuning, distilling, and self-hosting LLM models

Experience in traditional ML, with a focus on working with messy data and robust evaluation of model approaches

Proficiency with CI/CD, infrastructure as code, and containerization

What Success Looks Like

ML/AI models deployed with robust monitoring and significant customer impact

Agentic workflows improving internal/external operations

Infrastructure that is stable, observable, and automated

Successful iteration and delivery of new ML/AI products from concept to production

Ability to contribute to existing geospatial pipelines directly or through the use of AI