Powersoft19 has an immediate opening for a Machine Learning Scientist who will lead R&D efforts to deliver data-driven solutions to business problems. This individual must be knowledgeable in advanced mathematics to recognize patterns, identify opportunities, and make valuable discoveries leading to prototype development, model trouble shooting and product improvement. The successful candidate must be creative and willing to develop a thorough understanding of the science and business rationale behind various products, services, and methodologies.
Who you are: The Machine Learning Scientist is responsible for designing and supporting various AI based products/projects. You will be responsible for developing an expert-level understanding of the core scientific capabilities of different solutions, configuration of the system to maximize value, and delivery of analytical services to our customers. The Machine Learning Scientist must also be able to explain advanced concepts to business users and customer IT personnel.
- Apply knowledge of statistics, machine learning, programming, data modeling, simulation and advanced math to recognize patterns, identify opportunities, pose business questions and make valuable discoveries leading to prototyping development and product improvement
- Use a flexible, analytical approach to design, develop and evaluate predictive models and advanced algorithms that lead to optimal value extraction
- Generate and test hypotheses and analyze and interpret the results of product environment
- Create and implement mathematical models and scientific algorithms including the development of robust, rapid, and efficient numerical algorithms
- Execution of proof-of-concept pilots and implementation support
- Work with product engineers to translate prototypes into new products, services and features
- Provide guidelines for large-scale implementation
- Knowledge transfer throughout the organization, internal presentations, and white papers establishing thought leadership and capability excellence
- Coordinate and conduct testing and analysis of solutions across customer data sets
What You Have/Can Do as a Minimum:
- Advanced degree (Masters or Ph.D.) in Mathematics, Statistics, Engineering, CS, or Physical Sciences
- Strong background in statistical regression and modeling techniques (least squares, maximum likelihood, Bayesian estimation)
- Strong Python skills and familiarity with tools such as scikit-learn, pandas, etc.
- SQL and relational databases skills (MS SQL Server, Oracle, etc.)
- Knowledgeable of a broad range of mathematical techniques, tools, and modeling frameworks and able to assess their relative merits and applicability to specific problems
- Hands-on data analysis experience and the ability to produce data visualizations to present complex data graphically (distributions, scatter plots, sensitivity analyses)
- Ability to express real-world processes in the languages of mathematics and probability
- Excellent communications skills and a team player
What You Can Do to Stand Out:
- Proficiency with Cloud providers like AWS, GCP, Azure, etc.
- Proficiency with non-relational scalable data stores (Spark, Hadoop, MongoDB, etc.)
- Proficiency in an object-oriented language such as C++, Java, or C#
- Domain expertise in price optimization, demand forecasting, or inventory optimization
- Proficiency with data visualization libraries (Bokeh, Matplotlib, d3.js)
- Monte Carlo methods or other simulation techniques (Stan, PyMC, BUGS)
- Deep learning and autodifferentiation libraries (Tensorflow, Torch, Theano)
- Other optimization and machine learning techniques (linear/nonlinear programming, genetic algorithms, support vector machines, ensembling, etc.)
- Hierarchical and non-hierarchical clustering techniques (K-means, agglomerative clustering, divisive clustering, graph theory, QT clustering)
- Factor analysis, sensitivity analysis, spectral analysis, eigen systems analysis, Principal Components Analysis (PCA)
- Econometrics, Decision theory, discrete choice models, propensity modeling
- Reinforcement learning, active learning, dynamic systems