Position Overview

Blue Sky Innovators is seeking an undergraduate student or recent graduate to join our RF Machine Learning team as an RFML Product Workflow Engineer. This role offers a unique opportunity to architect and implement the complete data pipeline from raw RF signal ingestion through machine learning-based identification and fingerprinting. You'll work at the intersection of signal processing, machine learning, and software engineering to build production-grade systems that advance the state-of-the-art in RF domain analysis.

Responsibilities

Pipeline Architecture & Implementation

• Design and implement end-to-end data workflows from raw IQ data ingestion through RFML model inference
• Build robust data processing pipelines that handle various RF signal formats (SIGMF, raw IQ files, etc.)
• Architect scalable fingerprinting systems that extract and store unique RF signal characteristics
• Develop fingerprint matching and retrieval systems to identify previously observed signals
• Implement monitoring and logging throughout the pipeline to ensure data quality and system reliability

ML Integration & Optimization

• Integrate transformer-based and other neural network models into production workflows
• Optimize data preprocessing and feature extraction for ML model inputs
• Implement efficient batching and inference strategies for real-time or near-real-time processing
• Collaborate with ML researchers to transition models from development to production

Data Management & Storage

• Design database schemas for storing signal fingerprints and associated metadata
• Implement efficient similarity search and matching algorithms for fingerprint comparison
• Build APIs and interfaces for querying the fingerprint database
• Manage data versioning and provenance throughout the workflow

Software Engineering Best Practices

• Write clean, maintainable, and well-documented Python code
• Develop comprehensive unit and integration tests for pipeline components
• Use configuration management tools (e.g., Hydra) for flexible system parameterization
• Implement CI/CD practices for continuous deployment and testing
• Create visualization and reporting tools for pipeline performance metrics

Required Qualifications

• Currently pursuing or recently completed a BS/BA in Computer Science, Software Engineering, or related field
• Strong programming skills in Python with experience in software design patterns
• Understanding of data structures and algorithms
• Experience with version control (Git) and collaborative development workflows
• Ability to read technical documentation and learn new technologies independently
• Strong problem-solving skills and attention to detail
• Due to the nature of the work, you must be eligible for a security clearance and must be a U.S. Citizen.

Preferred Qualifications

• Coursework or projects in machine learning, signal processing, or data engineering
• Experience with scientific Python libraries (NumPy, SciPy, pandas)
• Familiarity with ML frameworks (PyTorch, TensorFlow)
• Knowledge of database systems (SQL, vector databases, or time-series databases)
• Experience with containerization (Docker) and orchestration tools
• Understanding of RF fundamentals or digital signal processing concepts
• Experience with package managers and virtual environments (conda, uv, pip)
• Familiarity with configuration management and experiment tracking tools

What You'll Learn

• Advanced RF signal processing and machine learning techniques
• Production ML system design and deployment
• Large-scale data pipeline architecture
• Domain expertise in RF fingerprinting and emitter identification
• Collaborative research and development in a cutting-edge technical domain

Work Environment

• Collaborative team environment working alongside RFML engineers and researchers
• Hands-on experience with state-of-the-art RF datasets and ML models
• Opportunities to contribute to novel research and development
• Flexible development environment with modern tooling (GitLab, VSCode, Jupyter, etc.)