SigOpt provides an enterprise, hosted service for model development and black box optimization. We're looking for brilliant and enthusiastic interns to learn, grow, and build with us!

Challenges:
SigOpt is a service that helps customers efficiently develop effective models in an efficient and data-secure fashion. As customers report the relationship between our recommended model designs and the associated performance, we implement algorithms internally to interpret this relationship and suggest new model designs which may reach higher performance. Such algorithms must be able to perform under various customer circumstances and computational restrictions, and they must yield robust performance for our customers to trust their model development to SigOpt.

Project Ownership: Interns are expected to apply a reasoned and rigorous judgment borne of a high level of maturity. Consequently, interns have responsibility over the scope and implementation of their projects, which may be critical pieces of our infrastructure and algorithms. We help interns develop their sense of how to power project progress through team communication and regular feedback.

Working with New Technologies: We provide support for interns to be able to quickly ramp up on the languages and platforms we use; we hope that learning these best practices will be a valuable at any company in the tech community.

Developing New Methodologies: Interns have the opportunity to explore new algorithms and ideas during their time here. Previous interns have worked on new concepts within Bayesian optimization, and adaptations of existing methodologies to suit customer needs. Furthermore, we hope interns bring their own strengths to the team and teach current team members new topics and ideas.

Communicating to the World: One of the key ways the research team contributes at SigOpt is by being effective communicators. This can be by documenting our work internally, developing feature/product documentation which speaks to advanced users, developing educational material to bolster SigOpt's conference presence, writing blog content to present ideas/methodologies to the broad community, or writing articles for conferences/workshops to the research community. Previous interns have worked on all of these and their contributions are still used daily to represent SigOpt's research.

User Specific Development: An important part of SigOpt's effectiveness lies in understanding our customer's needs. One option that interns can pursue is to work on understanding/reproducing customer use cases so that we can build tools/products which best serve them. This process is one part rigorous analysis, one part literature review, one part confirmation and a dash of judicious speculation. Some key insights in the past have come from this process, and we have numerous such opportunities in the near future.

Working at SigOpt:
Engineers are empowered to make big decisions about what they are working on and how they want to ship it. Interns work as core parts of the team: attending and presenting at meetings, scoping and documenting projects, regularly releasing production code. Past interns have, during their time at SigOpt, written articles, blog content, new features, new demos and some very important stuff about which we cannot speak. Many have gone on to great success at ICML, AAAI, NeurIPS (twice!), and full time within SigOpt itself!

Everyone on the research team has the opportunity to work independently with the support of the entire company. Interactions can range from standard corporate code reviews, to more grad school-style working sessions, to company-wide presentations to happy hour on Wednesday night.

As of 2021, SigOpt is a fully remote team under Intel.

Minimum Qualifications:

• Active Graduate student in Mathematics, Statistics, Computer Science, Operations Research, or related fields.
• 6+ months experience in Python
• Availability to a 3-month internship

Preferred Qualifications:

• Experience writing computational code
• Publications in Gaussian processes or Bayesian optimization
• Availability for a 6-12 month internship

This position is not eligible for Intel Immigration Sponsorship