Machine Learning Model Developer

Azura is seeking a full- or part-time skilled analyst to advance machine learning/Artificial Intelligence (ML/AI) models to support the Protected Species Division (Conservation Ecology Branch) of the NOAA Fisheries Northeast Fisheries Science Center (NEFSC). This role supports image-based line transect abundance estimates for marine mammals and sea turtles derived from imagery collected during crewed aerial surveys.  

Background 

The primary focus of this position is the development of a model to accurately identify individual animals in images recorded during aerial line transect surveys. The successful candidate will be responsible for training, refining, and testing ML/AI models for animal detection. This role also includes contributing to the development of the data management and processing workflow, starting from the point of data collection. Experience with VIAME, an open-source computer vision software platform for do-it-yourself artificial intelligence, is highly preferred, as images have already been annotated within this platform. Consideration may be given to candidates with experience in alternative model development platforms.  

Place of Performance and Work Schedule 

Work may be performed either onsite at the NEFSC laboratory in Woods Hole, Massachusetts, or remotely from within the United States. Work hours may be full-time (up to 40 hours per week) or part-time, depending on candidate availability. This position is expected to last approximately 6–11 months, with the possibility of an extension based on project needs and funding. During workdays, candidates must be able to respond to communications within 24 hours and participate in calls and meetings as needed during normal business hours. Travel is required (approximately 1-2 trips per year). This position does not offer visa sponsorship or relocation assistance. Estimated start date is late winter/early spring.

Duties 

• Develop and implement machine-learning models and/or algorithms to accurately identify and distinguish individual animals in images collected during aerial line transect surveys.
• Train, refine, and evaluate ML/AI models for animal detection and identification, including performance assessment and iterative improvement.
• Utilize existing annotated imagery within VIAME to support model development; candidates with experience in alternative computer vision or model-development platforms may also be considered.
• Contribute to the design and improvement of data management and image-processing workflows, beginning at the point of data collection and extending through model output and analysis.
• Conduct quality assurance and quality control (QA/QC) of imagery, annotations, and model outputs to ensure accuracy, consistency, and reproducibility.
• Document model development processes, assumptions, performance metrics, and limitations to support transparency and future use.
• Collaborate with scientists, analysts, and survey staff to ensure models align with survey objectives, data standards, and operational constraints.

 Required Qualifications 

• Bachelor’s degree in a relevant field of study (e.g., computer science, data science, engineering, biology, environmental science, or a related discipline).
• Strong attention to detail and demonstrated organizational skills.
• Experience developing, training, or applying machine-learning or computer-vision models for image-based analysis.
• Proficiency in R programming.
• Experience with database development and working with structured datasets.
• Familiarity with spatial plotting and data visualization tools.
• Ability to work effectively both independently and as part of a collaborative team.
• Eligibility to work in the United States.
• Ability to obtain and maintain a Public Trust clearance, which requires successfully passing a background investigation

 Preferred Qualifications 

• Experience with VIAME, an open-source computer vision platform for do-it-yourself artificial intelligence.
• Experience with alternative machine-learning or computer-vision model development platforms may be considered.

Thank you for applying. Azura Consulting LLC (Azura) is an Equal Opportunity Employer and does not discriminate on the basis of any status or condition protected by applicable federal or state law. We strongly encourage all qualified persons to apply for this position.
 
This recruitment process can take several weeks. Once we narrow the applicants, we conduct virtual interviews with chosen applicants and make a final selection. The chosen candidate then undergoes a federal background investigation which often takes 6 to 16 weeks; the candidate either passes or fails based on NOAA’s regulations. Contingent upon satisfactory completion of this investigation and United States work authorization verification, Azura will formally offer the position to the preferred applicant. If the applicant fails this investigation, the next applicant who meets the position’s criteria will be contacted. Azura will notify all other applicants if they are not chosen for the position.
 
About Azura

Azura is a woman-owned small business headquartered in the Dallas/Fort Worth area. We offer ‘big city’ resources with ‘small town’ customer service. Our primary consulting services are focused on natural resources, professional writing and editing, and educational outreach and communication. Please see our website for more information. www.azuraco.com