AI/ML Data Scientist
New York City (United States of America)
Open to multiple location: Global
GiveDirectly (GD) aims to reshape international giving – and millions of lives – by providing cash grants directly to the world’s poorest. The Brookings Institution estimates that $70B of cash transfers would be required to eliminate the poverty gap; the aid sector currently spends $135B each year, much of it on evidence-free interventions. GiveDirectly wants to change that, establishing cash transfers as a benchmark for foreign aid – as the index fund is a benchmark for the financial industry – and in the process accelerating the end of extreme poverty.
GD has raised over $500m since launching in 2011, delivered cash to more than 500,000 people, and launched offices in 7 countries. We’ve also supported large-scale, experimental research that indicates strong recipient impact of our programs. As a result, GD has been celebrated as one of the most innovative non-profit companies by Fast Company, while the growing cash transfer movement (and GD’s leading role within it) have been featured in the New York Times Magazine, This American Life, Foreign Affairs, and The Economist.
We’re looking for exceptional talent to grow our collaborative, diverse team. Joining GiveDirectly provides the opportunity to work alongside individuals who come from 21 different countries and speak 69 different languages. We’re actively working toward an equitable and inclusive environment for all team members, and seek candidates who will bring diverse perspectives and experiences to our organization. We recruit from organizations across all industries: our team has hailed from sectors including start-ups, government, consultancies, investment banks, and nonprofits.
Across our global offices, our culture is candid, analytical, non-hierarchical, and fast-paced. We value ability, adaptability, and willingness to learn. We offer competitive salary and benefits, as well as performance-based bonuses.
About the Role
GiveDirectly is investing in new, digital approaches to find, enroll and pay extremely poor populations at greater speed and scale. In partnership with experts at U.C. Berkeley and the Government of Togo, we have demonstrated a groundbreaking approach to identify and pay the extreme poor using AI/ML tools. Our work has been built off and contributes to a growing base of academic literature in this field, with our research partners recently publishing preliminary findings from our programs. Our AI/ML work has been featured in WIRED and Reuters.
We are looking for an expert who has experience developing cutting edge AI/ML technical solutions for the extreme poor. This individual will work with a data engineer to manage and analyze big data sources such as Call Detail Records (CDR) from telecommunications companies and high resolution satellite imagery. As GiveDirectly’s first data scientist, this individual will also help chart the course for our future technical investments in data science and be expected to support strategic decision making for new innovations. The role will report to the Director of Innovation and require the individual to travel and deploy for two to three months at a time.
- Design and direct data science investments that enable GiveDirectly to target, enroll and pay contactlessly at greater speed, scale and accuracy
- Manage collection of CDR and satellite data with MNOs and satellite companies
- Manage ground truth survey firm and sampling methodologies
- Build poverty prediction model/algorithm
- Produce analysis on algorithm performance covering metrics such as accuracy, recall, algorithmic bias etc.
- Support other GiveDirectly verticals such as field technology and fundraising generate data science insights
- Experience with storing and processing terabyte-scale datasets. Proficient in spark, dask, and/or other packages for distributed computing. Experience with SQL and database management.
- Deep understanding of ML algorithms. Ability to apply the appropriate methods and ability to benchmark and diagnose predictions to rapidly improve performance
- Experience developing an end-to-end ML pipeline with a scalable machine learning framework (TensorFlow, Torch, H2O, Spark MLib, scikit-learn)
- Experience with analysis methods for spatiotemporal data and social network data
- Experience with static and interactive visualization (Matplotlib, D3, ggplot, RShiny)
- Experience with quantitative research methods, including survey design and sampling methods
- Graduate degree in computer science, statistics, data science, or similar
- Ability to travel and deploy to international, on premise locations for up to 2 months at a time
- Strong communication skills and ability to compel technical and non-technical audiences
- Experience managing ground truth data collection firm with ability to manage sampling for bias and other quality metrics a plus
- Graduate coursework in development studies or development economics a plus
- Experience with satellite and call-detail record data a plus
- Working within the constraints of on-premise, low technical resource environment is a plus
- Secure server specification and maintenance. Managing a server cluster is a plus (likely done in conjunction with data engineer)
- Experience monitoring and maintaining data pipeline workflows (likely done in conjunction with data engineer)
In addition to the specific skills required for each role, we’re looking for candidates who are aligned with our GD Values, and who demonstrate our core competencies: emotional intelligence, problem solving, project management, follow-through, and fostering inclusivity.
**GD is committed to observing all local, national and international laws that protect children, vulnerable adults, and basic human rights of all. GD is committed to a policy of “zero tolerance for sexual exploitation, abuse, and harassment (SEAH)” and expects anyone who works for GD to uphold the protection and safeguarding of our recipients as a priority.**