AUGR LLC is seeking data scientists for Treasury FinCen data analysis. The qualified candidate will be responsible for collecting and analyzing large sets of structured and unstructured data from disparate sources. Additional responsibilities include: cleans and validates data to ensure accuracy, completeness, and uniformity, analyzes data to identify patterns and trends, devises and applies models and algorithms to mine the stores of big data, interprets data to discover solutions and opportunities.
Capabilities include R and Python; Standard machine learning packages such as panda and sklearn; Agile processes and procedures; Graph analytics experience, using software such as iGraph and NetworkX, and familiarity with common algorithms for community detection and node embeddings; Experience modeling tabular, unstructured, and semi-structed data as a graph; Experience with Graph databases (e.g. Neo4j) and relevant query and management skills; Experience with NLP, particular entity extraction. Experience using existing models and tools, such as Spacy, as well as training new models and evaluating their performance; Common machine learning algorithms and problem domains to include feature engineering, hyperparameter tuning, and performance optimization. Experience with SQL for advanced querying and joining of heterogeneous data sets.
Advanced candidates will additionally have statistics knowledge including common pitfalls in modeling and data analysis, sampling and data augmentation techniques, and mechanics of common ML algorithms and loss functions; Candidates with experience with existing FINCEN / DARPA analytic tools as well as experience with SAS and other COTS analytic tools highly preferred. Higher level candidates should have experience managing teams of staff conducting these efforts.
Requires a minimum Secret US Security Clearance; Top Secret preferred.