• Competitive
  • New York, NY, États-Unis
  • CDI, Plein-temps
  • Moody's
  • 2019-01-16

Machine Learning NLP/ Computer Vision Research Scientist / Principal Scientist

Lieu de travail : New York, NY, États-Unis

The Role / Responsibilities:
The Machine Learning NLP/ Computer Vision Research Scientist / Principal Scientist is a core member of the Emerging Business Unit (EBU) team, reporting to the Senior Director, ML and AI, Emerging Business Unit. This is a newly formed, highly visible team that is key to Moody's Analytics' (MA) long-term growth strategy that will lead our efforts to better understand and adapt to an environment characterized by widespread, technology-driven change. The EBU is charged with supporting innovation within our existing LOBs, developing opportunities in the "whitespace", enhancing our innovation process and understanding customers' technology preferences.

The Machine Learning NLP/ Computer Vision Research Scientist / Principal Scientist w ill, under supervision by the Senior Director, research, design, develop, and implement innovative Machine Learning, NLP, computer vision and deep learning solutions focused on natural language (text) and unstructured data that will advance Moodys Analytics capabilities across multiple business lines. On any day, the candidate could be doing any or all of the following:
  • Research emerging ML, computer vision (OCR), deep learning and NLP solutions applied to natural language (text) and unstructured data and be conversant with latest developments in these fields
  • Identify innovations in OCR, computer vision, text analytics and NLP leveraging ML/AI and deep learning to help advance automation, knowledge discovery, decision-making and insights, and streamline business processes or enable new capabilities in document understanding and knowledge extraction from text, images, speech and unstructured data,
  • Implement, prototype new algorithms for computer-vision OCR solutions, NLP, text analytics, and document understanding, using latest deep learning tools and framework (such as TensorFlow, Keras, Caffe) and develop, demonstrate, and validate novel ML, computer vision and NLP solutions in a hands-on role,
  • Evaluate custom, scalable deep learning and NLP solutions through prototyping, POCs and quantitative metrics, and handing off solutions to system engineers and stakeholder teams as needed
  • Develop new systems for evaluating NLP/NLU accuracy and performance and work with the data engineers in building better-annotated training corpora by assessing data collection and annotation processes
  • Discuss, suggest, and brainstorm new advanced technology solutions with team members
  • Explain complex models to non-experts, in layperson terminology to clients, stakeholders and managers, while also being able to discuss intricacies of complex algorithms with experts in the field
  • Prepare reports, presentations, for internal and external stakeholders, and as applicable, publish in peer-reviewed journals and magazines.
  • Attend, present at technical conferences, workshops, and meetups

  • Advanced or basic degree (recent PhD from top university / MS with 5+ years experience) in a quantitative field such as CS, EE, Information sciences, Statistics, Mathematics, Economics, Operations Research, or related, with focus on Machine Learning , AI , NLP , deep learning, and/or / data-driven statistical analysis & modelling
  • Demonstrated Experience in developing and applying ML, computer vision, NLP / deep learning & modeling solutions to specific problems involving OCR, NLP, text analytics, knowledge extraction and knowledge representation from text documents. Proven contributions to open-source ML/AI toolkits on Github, or developing custom products is desired.
  • Strong knowledge of the foundational theory and applications of machine learning, computer vision, deep learning, NLP, text analytics, unstructured data analytics, supervised/unsupervised learning. Experience with OCR is a plus.
  • Specific theoretical and applied programming experience with deep learning models such as LSTMs, RNNs, CNNs for NLP, OCR, document understanding, and text analytics.
  • Experience with NLP and machine-learning-based algorithms, solutions and frameworks for natural language understanding, document classification, entity recognition, document understanding, parameter extraction, and/or representation learning, as well as with the creation and evaluation of annotated training corpora.
  • Strong, proven programming skills (5+ years of programming experience) in Python, C/C++, Java, R , MATLAB, Scala, and with machine learning and deep learning and Big data frameworks including TensorFlow, Keras, Caffe, MXNet, Spark, Hadoop, and NLP frameworks such as NLTK, Stanford NLP, SpaCy, and others. 5+ years programming experience with writing complex programs and implementing custom algorithms in these and other environments.
  • Experience beyond using open source tools as-is, and writing custom code on top of, or in addition to, existing open source frameworks.
  • Proven capability in demonstrating successful advanced technology solutions (either prototype, POCs, well-cited research publications, open-source contributions, leadership in competitions from Kaggle, NIST, or others, and/or products) using ML, computer vision, NLP in one or more applications,
  • Additional experience with GPU-based training of deep learning models, and cloud environments such as AWS, Azure is a plus
  • Excellent communication skills (oral and written) to explain complex algorithms, solutions to stakeholders across multiple disciplines, and ability to work in a diverse team
  • Experience in an applied R&D environment , working in an agile, innovation-lab culture to bring cutting-edge technologies to fruition, from initial concept to implementation

Preferred .

Moody's is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, color, sex, gender, age, religion, national origin, citizen status, marital status, physical or mental disability, military or veteran status, sexual orientation, gender identity, gender expression, genetic information, or any other characteristic protected by law. Moody's also provides reasonable accommodation to qualified individuals with disabilities in accordance with applicable laws. If you need to inquire about a reasonable accommodation, or need assistance with completing the application process, please email accommodations@moodys.com.. This contact information is for accommodation requests only, and cannot be used to inquire about the status of applications.

For San Francisco positions, qualified applicants with criminal histories will be considered for employment consistent with the requirements of the San Francisco Fair Chance Ordinance. For New York City positions, qualified applicants with criminal histories will be considered for employment consistent with the requirements of the New York City Fair Chance Act. For all other applicants, qualified applicants with criminal histories will be considered for employment consistent with the requirements of applicable law.

Click here to view our full EEO policy statement. Click here for more information on your EEO rights under the law.

Candidates for Moody's Corporation may be asked to disclose securities holdings pursuant to Moody's Policy for Securities Trading and the requirements of the position. Employment is contingent upon compliance with the Policy, including remediation of positions in those holdings as necessary.