Job Description
*** Selected candidate must reside within two (2) hours of Client's Headquarters in Woodlawn, MD *** Selected candidate must be willing to work on-site at least 3 days a week. Position Description: - Provide support to the agency is creating a bias mitigation strategy and examine the Agency's race and ethnicity (R/E) data for consistency and completeness.
- Assess each of the identified legacy AI-systems for bias and develop the appropriate mitigation approach.
- Collaborate with the customer to remediate or rebuild the model for compliance.
- Compare the legacy and bias-mitigated models for performance, accuracy, and achievement of business objective.
- Refine and adjust the model until sufficient bias mitigation has been achieved and parity or acceptable performance with the legacy model has been achieved.
- Document findings and methods used, publish code to the repository, and prepare a standard operating procedure for future use cases.
- Present and communicate the intent of AI government policies and Executive Orders.
- Self-starter who can responsibly and independently advance work
- Work with client's leadership at all levels and present relevant information.
- All other duties as assigned or directed.
Key Required Skills: Experience with AI ethics, frameworks, and technologies. Requirements
Basic Qualifications - Bachelor's degree in Statistics, Applied Mathematics, Computer Science, or Information Science with 13+ years of experience
- Bachelor degree in Statistics, Applied Mathematics, Computer Science, or Information Science
- Minimum 8 Year (s) of Data Scientist experience
- Experience with IBM AI Fairness 360 (AIF360) - preference; Machine Learning (ML); Natural Language Processing (NLP); spaCy and NLTK, as well as language models like BERT and RoBERTa, are desired.
- Knowledge of TensorFlow, PyTorch, scikit-learn, NLTK, Azure ML (optional), Amazon Web Services EC2.
- Proven experience analyzing data, including data quality assessment, data governance, and data protection.
- Familiarity with ethical considerations related to AI, such as fairness, transparency, accountability, and bias mitigation.
- Experience with extracting insights from large datasets to determine bias.
- Excellent analytical skills to identify potential risks and propose effective solutions.
- Clear communication skills to convey complex ethical and technical concepts to various partners.
- Ability to collaborate with cross-functional teams.
- Providing problem solving skills, proven communication in written and verbal formats to various audiences to include executive leadership.
- Demonstrated an understanding of challenges in building production-worthy AI systems.
- Proactively seeking guidance, clarification and feedback.
- Must be able to obtain and maintain a Public Trust.
Required Skills - Strong knowledge of modern JavaScript and experience with ES6+ features, such as classes, arrow Demonstrating intimate understanding of privacy, data, and model governance activities including AI risks, oversight mechanisms, documentation, development processes.
- Knowledge of the legal and regulatory landscape surrounding AI, including privacy, ethics, liability, and intellectual property laws.
- Proven experience with the full system development lifecycle.
- Demonstrating an understanding of challenges in building production-worthy AI systems.
- Displaying an ability to manage senior stakeholders and client needs, including: building relationships with clients, approaching client in an organized manner, and delivering clear requests for information, facilitating alignment across multiple stakeholders;
- Displaying ability to research and analyze pertinent to the customer and technical matters.
- Demonstrating ability to contribute to a positive working environment by building relationships with team members;
- Collaborating with teams comprised of technical, data science, and business specialists.
- Collaborating with senior leaders and coordinating roll-out and change management for new processes.
Desired Skills - Master's Degree, Advanced Degree in Computer Science, other quantitative discipline.
- Proven experiences in AI ethics, governance, and compliance is a plus.
- Training and/or Certification in AI ethics or related fields is a plus.