Scientific Computing Engineer Intern
Sentrana Inc. - United States

This job posting is no longer available on Sentrana Inc.. Find similar jobs:Scientific Computing Engineer Intern jobs - Sentrana jobs

Job Opening: Scientific Computing Summer Internship
About Us Located in the heart of Washington DC, Sentrana provides companies with predictive analytics for better sales and marketing decisions. For the past five years, we have brought our capabilities to the foodservice industry generating optimal pricing and marketing recommendations covering tens of thousands of products and hundreds of thousands of customers on a daily basis. Through our enterprise application platform MarketMover®, we help wholesale distributors and suppliers put the terabytes of data sitting in their existing systems to work with advanced analytics and a family of applications that deliver predictive insight at the moment of truth – the instant the sale is being made.

Our application suite helps companies of all sizes accurately predict and optimize the impact of their sales and marketing activities, including pricing, promotion, product mix, and other related activities. The MarketMover software platform gives clients the ability to use our advanced analytics throughout the sales process and at the crucial moment of the sale. Our software as a service (SaaS) model allows our clients to get started quickly supported by our high performance computing infrastructure and allows us to continuously improve our software without costly deployments into clients’ internal IT environment.

Overview During your summer internship program within the Scientific Computing Group at Sentrana, you will develop your knowledge and skills for building reusable, high-performance scientific software. You will learn about development cost and computation cost as key contributors to total software, and see examples of the tradeoffs between them. You will build on this framework to arrive at a software development roadmap that attempts to optimize the development vs. computation cost. You will then apply this perspective to a real-world project in statistics and/or machine learning, with tangible results and deliverables within a 12-week span. Your deliverables will include not only a library of R/C/C++/CUDA code in Linux, but also a document summarizing your work as well as your learnings at the end of the project.

Work Plan A detailed work plan for your summer project depends on the specific problem that you will be working on, but at a high level, your internship will consist of three phases:
  • Phase I: Introduction (1 week) a. Day 1: Orientation
    b. Day 2: Philosophy and approach for building reusable, high-performance software
    c. Day 3: Programming environment and best practices for Sentrana Scientific Computing
    d. Day 4: Overview of the mathematics and algorithms relevant to the problem
    e. Day 5: Introduction to high-performance computing (OpenMP, MPI, CUDA)

  • 2- Phase II: Building Core Components (4-6 weeks) a. Week 1: Problem abstraction
    • Identify computation bottleneck for the problem)
    • Cast the computational bottleneck in the context of neighboring problems and existing software library)
    • Define a reusable abstraction of the computational component as well as list of client cases (i.e. higher-level algorithms that would utilize the computational component))

    b. Weeks 2-6: Implement and iterate
    • Implement and test the abstracted problem
    • Refine

  • 3- Phase III: Building Wrappers (4-6 weeks) a. Prioritize list of test cases from 2.a.iii
    b. Implement and test cases in the form of wrappers around the computational component developed in Phase II
  • Enrollment in a Ph.D. program in quantitative field such as Physics, Electrical Engineering, Statistics, Computer Science in a top-tier university
  • Undergraduate/graduate coursework in calculus, probability theory, and statistics
  • Experience writing scientific computing code in a low-level language such as C/C++/FORTRAN
  • Experience with parallelized code (OpenMP, MPI, CUDA) is preferred

Versatility and flexibility to adapt to new situations are essential. Ideal candidates have a GPA of at least 3.0 or higher on a 4.0 scale. All interns will work in our Washington DC office, and successful candidates may be offered full-time positions to begin after graduation.

Contact Us To apply, please submit your résumé and cover letter directly to us:
(Subject: Scientific Computing Intern)