Open Source Fellowship (Fall 2022)
almost 2 years ago
Paid fellowship
Philadelphia, PA, US
Philadelphia, PA, US
Job Description
Azavea's Cicero product is a database, API, and batch processing service that provides information about elected officials and legislative districts in the United States and several other countries. The current database includes more than 54,000 data elements, but this is only a fraction of the total number of democratically elected offices and elected representatives. Expanding and maintaining this database is a labor-intensive web research and data entry process. Azavea would like to accelerate this process using machine learning and natural language processing (NLP). In particular, we are seeking a fall intern to develop a prototype model that can automatically extract the name, district, legislative body, and contact information for legislators from official government webpages. This fellowship is open to anyone seeking to deepen their professional experience in the field of machine learning, and is not limited to students. We foresee the following plan for the fall:
• Complete a literature review on relevant techniques such as relation extraction and named entity recognition. • Write scripts to generate training and validation datasets from the Cicero database. • Implement and experiment with several models on the dataset using open source tools such as PyTorch, AllenNLP, Hugging Face Transformers, and/or spaCy. • Evaluate models and visualize results.
The code generated by this project will be made open-source, and we encourage publishing a blog post at the end of the fellowship. We are looking for candidates with experience in Python, data wrangling, deep learning (ideally using PyTorch), and (ideally) NLP. Azavea currently has three machine learning engineers who can serve as mentors for this project who all specialize in applying computer vision to geospatial and biomedical imagery. Therefore, we are looking for a strong, autodidactic intern who can work independently, and share their knowledge with the rest of the group. Interested in reading a blog post from a previous intern? Check out Transfer Learning from RGB to Multi-band Imagery.
The internship timeline is flexible. We are willing to start the project sooner or later to accommodate our intern's schedule. We are also willing to have the project last anywhere between 12 and 16 weeks. Additionally, the internship can be either hybrid (i.e. a mix of working from our Philadelphia office and remotely) or fully remote. Our office space in Philly would be open to the student 5 days a week if preferred.
A note on hiring during the COVID-19 pandemic Due to regulations in Philadelphia and our concern for the health and safety of our team, we expect to conduct most of the candidate interviews remotely. The majority of our colleagues generally work out of our Philadelphia office, but while our office has re-opened, most of us aren’t in the office full-time. Depending on the circumstances at the time of hiring, we are able to support remote on-boarding, including shipping relevant materials and a laptop to your home. We are fortunate to have invested in meaningful work-from-home tools and processes over the years, and have been able to continue providing a secure, flexible, and safe work environment for all of our colleagues. We ask for your patience as we adapt our hiring process as well, and are happy to answer any questions or concerns about the process.
Office and BenefitsOur Philadelphia headquarters is located in a brightly lit office on the 5th floor of a converted factory building in the Callowhill neighborhood, a short walk from Center City, the Reading Terminal Market, and SEPTA subway and regional rail stations. For bicyclists, we have in-house bike parking, showers, and lockers. The office itself is assembled as an open office plan with several smaller rooms for team meetings and concentration time.
• Work closely with experienced mentors. • Contribute to Azavea open source projects. • $9,000 stipend for 12 weeks, $12,000 stipend for 16 weeks. • Applicants that live outside the Philadelphia region will be eligible for assistance with relocation expenses on a discretionary basis. • Public Transit reimbursement: We encourage people to use public transit as an alternative to a car. We will reimburse you for your monthly public transit costs up to a maximum of $230/month. • Bicycle reimbursement: If you use a bike instead of walking or public transit, we will reimburse you for their monthly bicycle costs for commuting (including repairs, helmet, or purchase price of the bicycle) up to a maximum of $20/month. (Please note that due to IRS regulations, this benefit is mutually exclusive with the public transit benefit). • Flex Time options: Coordinate with your mentor if you need to come in late or leave early and make up the time when it works. • Three paid holidays - Thanksgiving, Christmas, and New Years • Paid Sick Leave: One day of paid sick leave is available, should you need it.
We welcome qualified candidates from all walks of life and value diversity in our company. We prohibit discrimination based on race, color, religion, ancestry, national origin, sex, sexual orientation, gender identity or expression, age, veteran status, military service, disability unrelated to job requirements, marital status, or domestic partner status.
• Complete a literature review on relevant techniques such as relation extraction and named entity recognition. • Write scripts to generate training and validation datasets from the Cicero database. • Implement and experiment with several models on the dataset using open source tools such as PyTorch, AllenNLP, Hugging Face Transformers, and/or spaCy. • Evaluate models and visualize results.
The code generated by this project will be made open-source, and we encourage publishing a blog post at the end of the fellowship. We are looking for candidates with experience in Python, data wrangling, deep learning (ideally using PyTorch), and (ideally) NLP. Azavea currently has three machine learning engineers who can serve as mentors for this project who all specialize in applying computer vision to geospatial and biomedical imagery. Therefore, we are looking for a strong, autodidactic intern who can work independently, and share their knowledge with the rest of the group. Interested in reading a blog post from a previous intern? Check out Transfer Learning from RGB to Multi-band Imagery.
The internship timeline is flexible. We are willing to start the project sooner or later to accommodate our intern's schedule. We are also willing to have the project last anywhere between 12 and 16 weeks. Additionally, the internship can be either hybrid (i.e. a mix of working from our Philadelphia office and remotely) or fully remote. Our office space in Philly would be open to the student 5 days a week if preferred.
A note on hiring during the COVID-19 pandemic Due to regulations in Philadelphia and our concern for the health and safety of our team, we expect to conduct most of the candidate interviews remotely. The majority of our colleagues generally work out of our Philadelphia office, but while our office has re-opened, most of us aren’t in the office full-time. Depending on the circumstances at the time of hiring, we are able to support remote on-boarding, including shipping relevant materials and a laptop to your home. We are fortunate to have invested in meaningful work-from-home tools and processes over the years, and have been able to continue providing a secure, flexible, and safe work environment for all of our colleagues. We ask for your patience as we adapt our hiring process as well, and are happy to answer any questions or concerns about the process.
Office and BenefitsOur Philadelphia headquarters is located in a brightly lit office on the 5th floor of a converted factory building in the Callowhill neighborhood, a short walk from Center City, the Reading Terminal Market, and SEPTA subway and regional rail stations. For bicyclists, we have in-house bike parking, showers, and lockers. The office itself is assembled as an open office plan with several smaller rooms for team meetings and concentration time.
• Work closely with experienced mentors. • Contribute to Azavea open source projects. • $9,000 stipend for 12 weeks, $12,000 stipend for 16 weeks. • Applicants that live outside the Philadelphia region will be eligible for assistance with relocation expenses on a discretionary basis. • Public Transit reimbursement: We encourage people to use public transit as an alternative to a car. We will reimburse you for your monthly public transit costs up to a maximum of $230/month. • Bicycle reimbursement: If you use a bike instead of walking or public transit, we will reimburse you for their monthly bicycle costs for commuting (including repairs, helmet, or purchase price of the bicycle) up to a maximum of $20/month. (Please note that due to IRS regulations, this benefit is mutually exclusive with the public transit benefit). • Flex Time options: Coordinate with your mentor if you need to come in late or leave early and make up the time when it works. • Three paid holidays - Thanksgiving, Christmas, and New Years • Paid Sick Leave: One day of paid sick leave is available, should you need it.
We welcome qualified candidates from all walks of life and value diversity in our company. We prohibit discrimination based on race, color, religion, ancestry, national origin, sex, sexual orientation, gender identity or expression, age, veteran status, military service, disability unrelated to job requirements, marital status, or domestic partner status.
Similar jobs
Mobilizing talent for a climate positive world