DARK WOLF SOLUTIONS, LLC — Department of Defense SBIR Phase I: A214-001
DARK WOLF SOLUTIONS, LLC — SBIR Phase I award from Department of Defense.
- Amount
- $223,738
- Agency
- Department of Defense · Army
- Program / Phase
- SBIR · Phase I
- Topic
- A214-001
- Solicitation
- 21.4
- NAICS
- —
- Place of performance
- VA
- Period
- 2022-08-15 → 2022-12-13
Description
Dark Wolf Solutions, LLC (Dark Wolf) proposes the Talent Acquisition aND Knowledge System (TANKS) solution to enable the Army’s Emerging Technology Leaders (ETL) Program to achieve human capital management efficiencies. Our approach capitalizes on our team’s proven successful Agile management approach, deep knowledge of a variety of Natural Language Processing (NLP) and clustering tools, and proven success with operationalizing Artificial Intelligence/Machine Learning (AI/ML) research. This approach consists of five steps: (1) Scope the Problem and Further Identify Requirements, (2) Conduct Data Discover, (3) Test and Analyze Results to Select the Optimal ML Model, and (4) Conduct Phase II Planning. Dark Wolf’s proposed TANKS solution will leverage one solution or several combined solutions, such as clustering, Latent Dirichlet Allocation (LDA), and deep learning, (in an ensemble approach) to provide the Army an easy-to-use tool to identify key skills, recruit and develop talent, and distribute talent across the enterprise. Our solution applies a clustering algorithm to build an ideal ETL profile of Knowledge, Skills, and Abilities (KSA). Each of those KSA’s is given a grade based on the average squared error within the cluster. TANKS then uses the KSA scores to build an overall KSA assessment for candidates where missing high grade skills are presented on the user’s profile as potential areas for development. (Talent Development.) In addition to the KSA score, TANKS will leverage deep learning to assess the narrative portions of a candidate's resume. The TANKS interface will aggregate these scores as well as their narrative score, enabling Army Human Resources professionals to identify highly rated candidates quickly and efficiently. These models, used in conjunction, will reduce overall hiring bias by building a score and profile without factoring in identifying characteristics of a candidate. (Skill Identification, Talent Recruitment, Talent Selection.) Finally, TANKS enables users to associate candidates with a team, and using the members’ individual scores, TANKS will designate an overall team score and skill assessment, identifying strengths and weaknesses. The TANKS teams’ profiles will allow the Army to track teams across the organization and optimally distribute talent. (Talent Distribution.)