Data scientist is a senior role that entails data analysis and code development, as well as project design and business development.
As a data analyst (applied mathematician) you will develop and test new methodologies for processing, analysing and interpreting large spatio-temporal datasets for the development and humanitarian sectors. You will report on your findings and methodologies in a way that is adapted to various stakeholders (academics, donors, end users).
As a developer, you will write/formulate queries to extract data from large/medium-size databases in a memory-efficient manner and will ensure sensitive data remain secure at all times. You will also support the development team in productifying/industrialising the methods/algorithms you develop.
As a senior team member, you will engage with end-users to understand their needs and identify which insights will be most relevant, as well as with donors to support business development.
Main tasks will include:
- Conducting exploratory and standard analysis of time series, geo-referenced time series, images and maps, and geo-referenced survey data
- Developing novel data analysis methods and algorithms, data metrics, statistical and mechanistic models
- Anomaly detection and data cleaning
- Data integration: combining multiple data sources in analyses (spatial data, survey data, mobile operator data)
- Assessing and quantifying data limitations and uncertainty of outputs
- Protecting subjects privacy, including acting in accordance with GDPR
Your code and SQL development tasks will include:
- Describing your algorithms and methods in a structured language (SQL) where appropriate.
- Creating actual SQL queries that mimic algorithms and methods.
- Revisiting SQL queries to make them more appropriate to particular data structures and data models available on different products.
- Revising SQL queries, that are intended to be run repeatedly or on large amounts of data, for speed and efficiency.
- Assessing the impact a data model has on how SQL needs to be written (often in a deconstructed form) in order to comply with a particular product architecture.
- Tailoring the process of running SQL queries (for example by running many smaller queries rather than one big query) to accommodate for either limited storage capabilities or processing power of target machines or to accommodate large data sets.
- Suggesting good data deletion strategies.
- Creating and documenting workflows of which the actual SQL may only be a part.
- Articulating which part of workflows requires what sort of (automatic or human) resource.
- Conforming to potentially restrictive architectures and interfaces by being creative with how those architectures and interfaces are used. This would include working with, and retrieving data from, APIs.
- Working on shared resources, managing your work so that it cannot prevent other operations from succeeding.
- Understanding the limitations appropriate to a data-sensitive environment.
In addition, you will also be involved in project design and business development, where tasks may include:
- Researching the needs of stakeholders and context
- Engage with end-users and donors
- Create analysis/research questions
- Conceptualise projects, activities and advise on the level of effort required
Finally, you will also be responsible for reporting analytical findings:
- Choose dissemination methods that are appropriate for users’ technical capacity
- Ensure results and conclusions are communicated clearly and can be understood by a non-technical audience
- Clearly communicate any limitations of the data
- Oversee and review data visualisations and reports
- Take into account ethical issues around how data will be used
- Disseminate work via presentations, workshops, conferences
- Reporting findings and progress on methods internally, to technical colleagues and non-technical staff
We seek talented and devoted data scientists/data analysts who look to apply their skills to support current humanitarian and development efforts, and seek to provide accessible and data-driven solutions to inform decision-making in LMICs. Our focus is to implement and operationalise the high-quality data products and solutions that we, and occasionally others, develop. The outputs of our work have the potential to be published in scientific journals. However, while we occasionally spend the necessary time to publish articles we do not do this systematically. Our focus is on operational, rather than academic impact.
In addition to scientific work, experienced candidates will have the chance to lead small to medium-sized projects or act as project director for certain initiatives.