Lead: Programme Analytics & Forecasting

m2m holds longitudinal, client-level data on a peer-led continuity-of-care model across multiple African countries. Today that data is used mainly to report what has already happened. The Lead:
Programme Analytics & Forecasting exists to make it say what is going to happen and what to do about it, by building, validating and maintaining predictive models and forecasts and translating them into decisions.

The role sits within Digital Products & Learning, between Strategic Information, which owns indicators and reporting, and the digital platform, where the data lives. Neither can produce forward-looking analysis alone; this role does, and explains it to the people who must act on it.

This is a senior individual-contributor role at Lead level, defined by advanced quantitative and predictive modelling capability. It is not a routine reporting, M&E coordination or data quality role; those sit with the Strategic Information function.

• Build, validate and maintain predictive models: client retention and disengagement risk, service demand forecasting, site and cohort performance projection, and caseload planning.
• Apply methods appropriate to the question, including time-series forecasting, risk scoring and classification, cohort and survival analysis, and geospatial analysis.
• State assumptions, quantify uncertainty, document limitations, and retire models that stop performing.
• Identify, unprompted, where a forward-looking question would serve the organisation better than the backward-looking one being asked.
• This area is ring-fenced at approximately 40% of the role's time and reviewed quarterly.

• Translate analysis into recommendations that country teams, programme leads and the Executive can act on, with a clear statement of what the analysis does and does not support.
• Present quantitative findings to non-technical audiences in plain language; this carries equal weight with technical capability.
• Provide analytical depth for donor reporting where needed, including results reporting requiring baseline, target and actual, and consortium-level analysis where m2m is Coordinating Party, working to the definitions owned by the Strategic Information Advisor.

• Specify what the analytics layer needs and work with the Senior Data Engineer on structure, quality and availability.
• Contribute to the data dictionary and indicator definitions, which the Strategic Information Advisor owns.
• Surface data quality problems with a proposed remedy, not only a complaint.

• Build analytical literacy across Programme, Impact & Delivery and country teams.
• Act as m2m's informed counterpart on any AI or machine-learning capability supplied by a vendor or the enterprise AI function: able to interrogate how a model works, what it was trained on and where it fails.
• Advise on data protection and ethical use of client-level health data in analytics, including data on adolescents and young people.

• Qualification. Master's degree (or higher) in data science, statistics, biostatistics, epidemiology, applied mathematics or a related quantitative discipline.
• Experience. At least ten years applying quantitative methods to real operational data, including at least three years building, validating and deploying predictive or forecasting models whose outputs were used in operational decisions, and at least two years in health, development or the public sector.
• Predictive, not only descriptive. Able to describe a model they built, what it predicted, how it was validated and what decision changed as a result. Experience limited to descriptive reporting,…