Artificial Intelligence / Machine Learning Engineer (Clearance Required)

Job Number: 1509

Riverside Research is seeking an Artificial Intelligence / Machine Learning Engineer to support existing contracts to prototype and develop automation solutions to NASIC’s most difficult Scientific & Technical Intelligence problems. As a highly valued and sought-after Riverside Research employee, you will be part of a highly skilled and integrated team that analyzes intelligence data to discover opportunities for automated solutions.

All Riverside Research opportunities require US citizenship.

Job Responsibilities:

AI/ML algorithm development for prototype applications in remote sensing:

• • Develop prototype AI/ML algorithms and associated software tools using Python and the Python/TensorFlow API
  • Train AI/ML models and tune their hyperparameters for a given dataset and algorithm objectives
  • Visualize hyperparameter optimization spaces with Tensorboard for selection of optimal parameters for a given parametric (functional) TensorFlow model

AI/ML dataset generation, curation, and management:

• • Provide customized solutions to data quality control that ensure accurate functional mappings for AI/ML algorithms on complex remote sensing datasets
  • Develop databases / data lakes / data warehouses for organizing both structured and unstructured datasets

AI/ML algorithm R&D:

• • Apply machine learning and general computer vision best practices and methods to analyze and exploit large, complex remote sensing datasets from a variety of remote sensing phenomenology
  • Keep up with the SoTA practices for AI/ML, perform relevant R&D, and implement new and innovative ideas in machine learning and high-performance computing to solve long-standing remote sensing “big-data” exploitation problems

Software development, documentation, and coding best practices:

• • Contribute and adhere to the AI team’s standards for reviewing and unit-testing code, lead or participate in team-wide code reviews, and adhere to standardized documentation practices
  • Utilize Python PEP8 standards

Required Qualifications:

• Must have minimum of Secret with able to obtain and maintain a TS/SCI clearance.
• Bachelor’s Degree in either Electrical Engineering, Mathematics, Statistics, Physics, Computer Science, or related field of study
• Must demonstrate proficiency in Python-based end-to-end AI/ML model development lifecycle using a recent deep learning platform (TensorFlow preferred)
• Awareness of version control, branches, merge conflict resolution, and git in general
• Proficient in collaborative Office 365 tools such as MS Word, Excel, and PowerPoint
• Ability to work closely with subject-matter experts to develop tools, algorithms, and datasets needed for developing relevant and useful AI/ML prototype algorithms
• Self-driven, strong analytic, inferencing, critical thinking, and creative problem-solving skills
• Communicates highly technical results and methods clearly and succinctly

Desired Qualifications:

• Advanced degree (MS/PhD) in Data Science, Mathematics, Statistics, Computer Science, a Physical Science or Engineering is strongly desired
• Active TS/SCI Security Clearance
• Experience with DoD intelligence production processes and workflows
• 3 years operational experience in radar signal processing analysis, overhead imagery analysis, orbital mechanics, and/or electronic warfare data analysis
• 2 years experience using data visualization tools and libraries in Python
• Visualizations/Web Development Skills (e.g., Tableau, MEAN stack - MongoDB, ExpressJS, AngularJS, NodeJS)
• Experience with large (1 GB ) image data and formats such as HDF5, JSON, GEOTIFF, TFRecords, etc.
• Experience in development of distributed, web-based systems, service-oriented architectures, front-end user interfaces, and back-end databases are a plus
• Experience with interpretability of deep learning computer vision models including visualization and reasoning about model latent spaces and activation maps to assess model effectiveness / weaknesses
• Familiarity in differences of supervised learning vs. unsupervised learning techniques