Quantitative Analytics Specialist

RESPONSIBILITIES:

  • Design, develop, and deploy AI/ML models using state of the art techniques available in the open stack (Python/PySpark/PyTorch) and/or vendor solutions
  • Partner with LOB leads to frame the problem, explore various ML/DL model architectures and methodologies, generate required artifacts related to model development life cycle (MDLC), author the model development document, and deliver AI models that meet business needs
  • Adhere to corporate model risk policy and ensure compliance with model risk management
  • Working with other data science teams to identify, gather, retain, and publicize modeling artifacts required for approved and repeatable processes
  • Work with AI technology and production teams to operationalize models
  • Work effectively in an agile project management methodologies for data science
  • Knowledge sharing with members of the team and across the organization on topics including machine learning algorithms, hyper-parameter tuning/search, and traversing across multiple big data platforms
  • Contribute to the CoE data science team’s group effort to stay current with the cutting edge NLP/ML/DL algorithms, methodologies in the open source community and vendor solutions.

Required Qualifications

  • 4+ years of experience in an advanced scientific or mathematical field
  • A master’s degree or higher in a quantitative field such as mathematics, statistics, engineering, physics, economics, or computer science
  • 4+ years of statistical modeling experience
  • 3+ years of Python experience
  • 2 + years of experience using quantitative machine learning techniques

Desired Qualifications

  • 2+ years of Hadoop experience
  • 1+ years of Natural Language Processing (NLP) experience

Other Desired Qualifications

  • Hands on familiarity with machine learning and statistical modeling techniques applying machine learning techniques such as neural networks, random forest, GBM and SVM, Probabilistic Graphical Models, Deep Learning architectures, using open-source languages like Python, PySpark and/or PyTorch.
  • Experience with model monitoring and performance tracking
  • Hands on experience with deep learning toolkits such as Tensorflow, Keras, PyTorch, Dynet
  • Hands on experience writing data processing and data pipeline for model development including gathering and building datasets to collect intents, joining and aggregating source datasets, cleaning messy data, designing feedback loop on data needs
  • Experience building intent recognition and classification models. Experience with phrase level identification. Experience with language models such as BERT, ELMO, CRF
  • Experience with active learning and reinforcement learning
  • Experience with AI model transparency and explainability studies
  • Strong acumen in diagnosing and resolving data issues
  • Exceptional analytical, critical thinking, quantitative reasoning skills, and problem-solving skills. 

Analytic Consultant

Responsibilities:

  • Lead and/or support development of hierarchical categorization taxonomies using NLP unsupervised/supervised methodologies including enrichment of unstructured data to derive themes and trends, with specific focus on early identification of emerging themes.
  • Develop non-model and model development to generate actionable insights and uncover patterns (structured or unstructured) from multiple sources, while working in partnership with teams across Conduct Management.
  • Perform text analysis and advanced analytics including using Natural Language Processing, Machine Learning techniques and programming languages to derive relevant analysis and metrics, including building proof of concepts to determine value of implementing in future projects.
  • Collaborate with team to develop data driven analytics capabilities to identify anomalies, outliers, emerging trends and conduct risks across the Enterprise. Primary areas of focus include Internal Fraud, Sales Practice, Ethics, and allegations
  • Collaborate with team to design complex projects/analyses generally spanning multiple products/business lines/functional areas/data sources including analytical application of industry-leading experimental techniques or exploratory data analytics to support teams across Conduct Management.
  • Partner with other Conduct Management and LOB stakeholders to develop, and enhance analytical capabilities related to Conduct Management’s first line of defense
  • Offer strategic recommendations to senior stakeholders and leadership based on thought-provoking and sound analytics including communicating technical and business details to ensure understanding.

Required Qualifications

  • 6+ years of experience in one or a combination of the following: reporting, analytics, or modeling; or a Masters degree or higher in a quantitative field such as applied math, statistics, engineering, physics, accounting, finance, economics, econometrics, computer sciences, or business/social and behavioral sciences with a quantitative emphasis and 4+ years of experience in one or a combination of the following: reporting, analytics, or modeling
  • 2+ years of Python experience
  • 2+ years of Natural Language Processing (NLP) experience
  • 2+ years of statistical modeling experience

Desired Qualifications

  • Extensive knowledge and understanding of research and analysis
  • Strong analytical skills with high attention to detail and accuracy
  • Excellent verbal, written, and interpersonal communication skills
  • Experience with design, implementation and governance with Artificial Intelligence, Natural Language Processing or Machine Learning Architecture
  • Knowledge and understanding of analytical methods used in: statistics or predictive modeling methods (regression modeling, clustering, pattern recognition, graphics)
  • 2+ years of Tableau experience

Other Desired Qualifications

  • Model governance experience including writing model development documentation
  • Developing models to identify/detect emerging themes
  • Preprocessing (tokenization, part-of-speech tagging, parsing, stemming)
  • Semantic analysis (named entity recognition, sentiment analysis)
  • Modeling and word embeddings/representations (RNN / ConvNets, TF-IDF, LDA, Word2Vec)
  • Working with data science analytic tools (Watson Studio, Watson Discovery, IBM SPSS Modeler, Data Robot)
  • Experience with Tableau for data visualization
  • Expert ability to create compelling presentations in Power Point for executive level audiences
  • Experience in fraud, BSA/AML, investigations of financial crimes transactions or policy violations, risk management or compliance
  • Analytical and detail oriented with the ability to prioritize, execute and deliver projects on time
  • Strong desire and ability to independently solve problems and self-prioritize multiple tasks

Job Expectations

  • Ability to travel up to 10% of the time