Data & ML Infrastructure Engineer

As a Data & ML Infrastructure Engineer, you will build the data infrastructure that enables HavocAI to develop, evaluate, and continuously improve autonomous systems.

You will own the pipelines and tooling that transform large volumes of video, imagery, telemetry, sensor data, autonomy logs, and mission data into organized, searchable, and reproducible datasets. Your work will provide Autonomy and Perception engineers with the high-quality data they need to train models, evaluate system performance, reproduce failures, and improve deployed capabilities.

A major focus of this role will be HavocAI’s internal video and telemetry data lake, including ingestion, storage, indexing, metadata, curation, quality, labeling, and dataset generation.

This is a hands-on engineering role for someone who enjoys building scalable infrastructure and turning messy real-world data into reliable engineering tools and ML-ready datasets.

Data Infrastructure & Pipelines

• Build and maintain infrastructure for video, imagery, telemetry, sensor data, autonomy logs, mission data, and field-test data.

• Own data ingestion, storage, indexing, metadata, access patterns, and lifecycle management within HavocAI’s data lake.

• Develop scalable pipelines that transform raw operational data into curated datasets for ML training, evaluation, debugging, and analysis.

• Build tools for searching, filtering, tagging, and retrieving data across platforms, missions, operating conditions, and events.

• Design infrastructure capable of handling large volumes of multimodal operational data efficiently and reliably.

Dataset Curation & ML Enablement

• Build workflows to select, clean, label, validate, and version datasets.

• Partner with Autonomy, Perception, Software, and Field Operations teams to identify high-value data for model development and system evaluation.

• Support annotation and labeling workflows for video, imagery, tracks, telemetry, and other ML inputs.

• Develop reproducible dataset-generation workflows for training, validation, regression testing, and benchmarking.

• Integrate datasets and data infrastructure with model training, experiment tracking, evaluation, and deployment workflows.

• Support multimodal dataset construction, including synchronization and alignment across sensors and data streams.

Data Quality & Reliability

• Develop automated checks for missing streams, corrupted files, synchronization issues, metadata gaps, labeling errors, and pipeline failures.

• Establish standards for dataset quality, lineage, versioning, and reproducibility.

• Build monitoring and observability around critical data pipelines and infrastructure.

• Troubleshoot complex data and infrastructure issues and drive them through resolution.

• Use field data, logs, and test results to help engineering teams understand system performance and identify opportunities for improvement.

Developer Tools & Collaboration

• Build self-service tools that make operational data easier for engineers to discover, access, analyze, and use.

• Partner closely with Autonomy, Perception, Software, Simulation, Field Operations, and Program teams.

• Translate engineering and ML requirements into scalable data capabilities.

• Improve workflows for replaying, visualizing, analyzing, and comparing operational data.

• Maintain clear documentation, data standards, and best practices for internal data use, governance, and security.

What We’re Looking For

• A Bachelor’s degree in Computer Science, Data Science, Machine Learning, Electrical Engineering, Computer Engineering, Robotics, Applied Mathematics, or a related technical field.

• 3+ years of experience in data engineering, ML infrastructure, data platforms, backend systems, MLOps, or related engineering roles.

• Experience designing and operating production data pipelines for large-scale structured, semi-structured, or unstructured datasets.

• Experience working with video, imagery, time-series telemetry, sensor data, logs, or other high-volume operational data.

• Strong programming skills in Python and SQL.

• Experience with cloud storage, object stores, data lakes, databases, distributed processing, or modern data platforms.

• Familiarity with dataset versioning, metadata management, data lineage, access controls, and reproducible data workflows.

• Strong software engineering fundamentals, including testing, reliability, maintainability, and observability.

• Strong debugging skills and comfort working across complex data pipelines and production infrastructure.

• Ability to operate independently and take ownership in a fast-moving engineering environment.

• U.S. citizenship and ability to obtain and maintain a U.S. Government security clearance.

Nice to Have

• Experience with ML infrastructure, MLOps, training pipelines, experiment tracking, model evaluation, or model registries.

• Experience managing video, perception, telemetry, or autonomous-system datasets.

• Experience with technologies such as S3-compatible storage, PostgreSQL, Spark, Ray, Airflow, Dagster, Kubernetes, Docker, or Kafka.

• Experience with data catalogs, dataset versioning platforms, feature stores, or labeling tools.

• Experience building search, replay, visualization, or analysis tools for video, telemetry, logs, or sensor data.

• Experience supporting annotation workflows for computer vision, perception, tracking, or autonomy.

• Familiarity with sensor synchronization, timestamp alignment, calibration metadata, log replay, or multimodal dataset construction.

• Experience with security, access controls, auditability, and data-handling requirements in government or defense environments.

• Experience supporting defense, robotics, autonomy, aerospace, or dual-use technology programs.

• Active or prior security clearance.

What Success Looks Like

Within your first 12 months, you will have:

• Built reliable pipelines that move operational data from field capture into organized and searchable storage.

• Made HavocAI’s video, telemetry, and sensor data significantly easier for engineers to discover and use.

• Established reproducible workflows for creating high-quality datasets for model training, evaluation, and regression testing.

• Improved data quality, lineage, metadata, and observability across critical pipelines.

• Enabled Autonomy and Perception teams to move more quickly from field data → insight → dataset → model improvement → deployment.