Daniel Perrefort

Astro-physicist turned software developer. I focus on training scientific research teams on new technologies for developing and deploying software. I also oversee the creation of innovative software tools used to increase the efficiency and effectiveness of my fellow team members and work within the Dark Energy Science Collaboration (DESC) to design scalable data analysis pipelines run at the NERSC supercomputer center.

Current Affiliations

  • Faculty Consultant: University of Pittsburgh, Center for Research Computing
  • Research Assistant Professor: University of Pittsburgh, Physics Department
  • Collaboration Member: Dark Energy Science Collaboration

Experience

Data Visualization

Expert in using data-driven communication to collaborate effectively with teams having diverse levels of experience.

Design and Delivery

Extensive experience developing software solutions to meet detailed project requirements and satisfy complex design conditions.

Test Driven Development

Well versed in writing and debugging comprehensive test suites using multiple testing frameworks.

Continuous Integration

Frequently works in large collaborative settings using continuous integration/deployment services to streamline provided services.

Diverse Project Experience

Experience developing in diverse software ecosystems to build web, API, and command-line based software tools.

Cloud Based Deployment

Familiar with project deployment to Google Cloud services like App Engine, Big Query, and Cloud SQL.

Portfolio

Egon is a light-weight framework for building parallel data analysis pipelines in Python by combining modular design principles with automated pipeline validation tasks.

The Supernova Atmospheric Simulation package (snat_sim) is used to support research into the impact of atmospheric calibration techniques on astronomical observations.

PG-Broker is a cloud-base alert broker that handles the real-time data processing for large-scale astronomical surveys like the Legacy Survey of SPace and Time.

SNData is a scalable Python API for accessing data releases published by astronomical surveys. It supports the development of flexible analysis pipelines by mapping complex data models onto an intuitive and easy to use interface.

pwv_kpno is a scientific Python package that provides models for the atmospheric transmission due to precipitable water vapor (PWV) at user specified sites.

Resume

Summary

Daniel Perrefort

Expert in data driven communication with over 6 years experience in software development and pipeline design. Extensive experience working in a collaborative setting to meet detailed project requirements.

Employment

Research Assistant Professor

2021 - Present

University of Pittsburgh, Pittsburgh, PA

  • Collaborate with team members to design and implement DevOps workflows for increased project stability.
  • Assist in the standardization of operating procedures to streamline daily workflows.
  • Provide consultation on modern software design and development practices.
  • Plan and execute training seminars on fundamental software development principles such as version control, continuous integration, and continuous deployment.

Education

Ph.D. of Astrophysics and Cosmology

2015 - 2021

University of Pittsburgh, Pittsburgh, PA
Thesis: Building a Better Candle: The Calibration and Classification of Type Ia Supernovae in the Upcoming Legacy Survey of Space and Time

Master of Physics

2015 - 2017

University of Pittsburgh, Pittsburgh PA
Focus: Physics and Astronomy

Bachelor of Physics

2011 - 2014

University of Connecticut, Storrs CT
Focus: Applied Physics

Service Positions

Dietrich School of Arts & Sciences Graduate Council

2018 - 2020

Committee Member

  • Voted on university policy including the approval of curriculum and instructional programs, admissions issues, and degree requirements.
  • Assisted in the review of TA supervision and training practices in multiple academic departments.
  • Provided insight into new graduate student programs targeted at improving placement opportunities, time-to-degree, and graduate student support.

Dietrich School Graduate Student Organization (GSO)

2017 - 2020

Physics Department Representative

  • Coordinated cross-department communication between multiple teams of representatives to facilitate the sharing of solutions to common logistical challenges.
  • Chaired the Graduate Travel Grant Committee to provide travel funding in support of graduate student lead research projects.

Association of Physics and Astronomy Graduate Students

2017 - 2018

Co-President

  • Leveraged a university allocated budget to maximize the number of services provided to physics graduate students.
  • Coordinated with multiple university offices to plan and execute multiple yearly training events.
  • Organize interviews between students and prospective faculty during department hiring initiatives and present student feedback to the department hiring board.
  • Work closely with faculty to plan and execute recruitment events for prospective students.

Talks

Some of the more recent talks I've given. For a full list of my talks and publications, please see my CV.

    • Aug 11, 2020 Rubin Observatory Project and Community Workshop, Virtual
    • Identifying 91bg-like Supernovae with Rubin-LSST
    • June 3, 2020 Pitt Summer Bootcamp, Virtual
    • Good Coding Practices
    • Apr 3, 2020 Astro-Lunch Seminar Series, University of Pittsburgh
    • Supernova Classification: How to Find What You’re Looking For
    • Feb 4, 2020 Astro-Snacks Workshop, University of Pittsburgh
    • A Crash Course in Software Verification and Unit Tests
    • Jan 21, 2020 DESC Collaboration Meeting, University of Arizona
    • Nov 20, 2020 Astronomy on Tap
    • Things that go Bump in the Night: How to find space stuff
    • Sep 26, 2019 Midwest Workshop on Supernovae and Transients, Ohio State University
    • Photometric 91bg Classification using Light-Curve Morphology
    • Feb 4, 2019 Astro-Lunch Seminar Series, University of Connecticut
    • Exploring LSST calibration strategies with GPS satellites and atmospheric modeling
    • Oct 4, 2018 LSST/DESC Photometric Calibration Workshop, LPNHE, Paris France
    • pwv_kpno: A Python Package for Modeling PWV transmission using SuomiNet GPS
    • Sep 18, 2018 Astro-Hacks Programming Series, CMU (Invited)
    • How to Write Code Better and Faster: An Introduction to Using IDE's in Your Daily Workflow

Travel

I get to work with an amazing team of individuals from around the globe. Here are a few recent workshops I've attended along with their official workshop photos.

2021

  • August: Rubin Observatory Project and Community Workshop
  • July: Dark Energy Science Collaboration Bi-Annual Meeting
  • Apr: LSSTC Enabling Science Broker Workshop - Part II
  • Feb: Dark Energy Science Collaboration Bi-Annual Meeting

2020

  • Oct: LSSTC Enabling Science Broker Workshop - Part I
  • August: Rubin Observatory Project and Community Workshop
  • July: Dark Energy Science Collaboration Bi-Annual Meeting
  • Jan: Dark Energy Science Collaboration Meeting

2019

  • Sep: TOM Toolkit Workshop
  • Sep: Midwest Workshop on Supernovae and Transients
  • Aug: LSST Project and Community Workshop
  • Jul: Coordinated Theoretical-Experimental Project on QCD
  • Jul: Dark Energy Science Collaboration Meeting
  • May: PITT-PACC Phenomenology Symposium
  • Feb: Dark Energy Science Collaboration Meeting
  • Feb: Dark Energy Science Collaboration Broker Workshop
  • Feb: UConn Astro-Lunch Seminar Series
  • Jan: The Future of SN Host Galaxy Studies

2018

  • Oct: LSST / DESC Calibration Workshop
  • Sep: Phipps Science Communication Workshop
  • Jul: Dark Energy Science Collaboration Meeting
  • Jun: Data Visualization and Exploration in the LSST Era
  • May: PITT-PACC Phenomenology Symposium
  • May: LSST Data Science Fellowship Workshop
  • Apr: New Advances in NIR Type Ia Supernova Science
  • Feb: Dark Energy Science Collaboration Meeting

2017

  • Jul: Sci-Coder Development Workshop
  • Jun: Summer School in Statistics for Astronomers
  • May: PITT-PACC Phenomenology Symposium