Small-Area Estimation & Data Fusion · Survey Methodology · Statistical Software
Twenty years of small-area estimation, survey methodology, and official statistics: from the WHO/UNICEF immunization estimates for 195 countries to county-level health estimates across all 102 Illinois counties, backed by the svy ecosystem and svyLab, the open-source tooling that makes the numbers reproducible.
import svy design = svy.Design(stratum="sdmvstra", psu="sdmvpsu", wgt="wtmec2yr") sample = svy.Sample(data=nhanes, design=design) # mean SBP among adults with diabetes, by age group sbp = sample.estimation.mean( y="sys_bp", by="age_group", domain="diabetes == 1", )
About
I'm Mamadou S. Diallo, a Ph.D. statistician based in Hoboken, NJ. For two decades I've turned sparse, multi-source data into defensible local estimates: small-area estimation, survey methodology, and the integration of survey, administrative, census, and geospatial data, across immunization, HIV, chronic disease, and official statistics.
Lead sampling statistician on the Population-based HIV Impact Assessment surveys (8 countries, with CDC/PEPFAR). Designed the NHANES 2012 compositing strategy at Westat, directed the WHO/UNICEF immunization estimates (WUENIC) for 195 countries at UNICEF, and developed sampling methodology for Statistics Canada's national health surveillance system (CCHS) earlier in my career. Author of Samplics (270K+ downloads, JOSS) and architect of the open-source svy ecosystem.
I also build the infrastructure evidence runs on. Solo architect of svyLab, a multi-tenant analytics platform with architectural sandbox isolation, lifecycle and classification governance, AI-assisted analysis with full provenance, and a 327-test invariant suite. Built so that survey-correct estimation is reproducible and auditable by construction: every result carries its model, code, weights, and CI.
My long-term goal is to modernize how survey data gets analyzed: open, reproducible, survey-correct tooling as a genuine alternative to SAS, SPSS, and Stata. The svy ecosystem and svyLab are open source and a career-long commitment — I maintain them the way statisticians maintain CRAN packages, alongside any role. Currently open to senior positions in survey methodology, small-area estimation, and population-health measurement.
Ph.D. Statistics, Carleton University
(advisor: J.N.K. Rao)
M.Sc. Statistics, Université Laval
ASA (since 2010) · AAPOR
Guest Editor, JSSAM 2025
English & French: fluent in both speaking and writing
Hoboken, NJ: open to remote and on-site engagements
Currently focused on small-area estimation, data fusion, and the svy ecosystem.
Services
I help health agencies, statistics offices, foundations, and research institutes turn complex, multi-source data into decision-grade local estimates, and build the software that makes those estimates reproducible.
Reliable subnational estimates when direct samples are too small: Fay–Herriot, unit-level, and spatial Bayesian models. Refined through years of applied work with national statistics offices and through the Healthy Illinois Analytics platform (NORC, all 102 counties).
Harmonizing survey, administrative, census, and geospatial data into decision-ready estimates: the discipline I built at WHO/UNICEF for 195 countries. Fit-for-purpose data assessment, discrepancy resolution, and reporting under quality frameworks (GATHER, AAPOR).
End-to-end design of complex multi-stage household surveys: frame development, sample size, stratification, cluster selection, weighting, calibration, variance estimation. Compliant with international standards. PHIA, NHANES, MICS, DHS, LSMS, and custom programs.
Production-grade analytics with multi-tenant isolation, AI-assisted analysis with provenance, lifecycle governance, and reproducibility primitives: the infrastructure any serious estimation program needs but rarely builds well. Architect of svyLab and the open-source svy ecosystem.
Long-term technical backstopping for national statistics offices and research teams: sampling design, weighting, small area estimation, reproducible analysis. Delivered in English or French. Recent: NBS Tanzania (World Bank, 2026), INSBU Burundi (2025), Ethiopia, Senegal, Vietnam.
Extending design-based survey methods to causal and transportability questions: survey-weighted IPTW and doubly robust estimation (svy-causal), validated against NHANES. Methods paper in preparation.
Selected work
A selection of the platforms I've built, the methods I've developed, and the population-based studies that ground both.
Multi-tenant analytics platform for health-data evidence generation
Sole architect and developer. Architectural sandbox isolation (cross-org reads return 404 by design, not just by access control), lifecycle and classification governance with audit logs, AI-assisted analysis where every output persists with full provenance: model, prompt, code, tokens, cost, timestamp. 327 passing tests including structural CI invariants, cross-org sandbox enforcement, existence-leak contracts, and quota correctness. Built so that survey-correct estimation is reproducible and auditable by construction: every number can be re-run from its inputs.
At NORC, led development of the Healthy Illinois Analytics Platform, a Python-based platform producing county- and community-level health estimates across all 102 Illinois counties using small area estimation. Integrated American Community Survey administrative data with survey microdata under disclosure control.
At UNICEF, led the annual production of WHO/UNICEF immunization coverage estimates for 14 vaccines across all 195 countries. Integrated administrative reporting (DHIS2), household-survey microdata (DHS, MICS), and programme surveillance under GATHER reporting standards, with fit-for-purpose assessments across data modalities.
Production-grade Python library for sample selection, weighting, estimation, and small area estimation. Published in the Journal of Open-Source Software (JOSS, 2021). The reference implementation for survey statistics in Python.
JOSS paper →At Westat, developed model-based small area estimation for state-level crime rates using 15 years of National Crime Victimization Survey data. Published the R package sae2 on CRAN, used for federal subnational statistics.
Lead sampling statistician at Westat for the Population-based HIV Impact Assessment surveys across Cameroon, Côte d'Ivoire, Malawi, Namibia, Tanzania, Uganda, Zambia, Zimbabwe, in collaboration with ICAP at Columbia and CDC/PEPFAR. Multi-stage probability samples for HIV prevalence, incidence, viral-load suppression, and ART coverage at national and subnational levels.
Survey-weighted causal inference: IPTW, stabilized weights, and doubly robust estimation extending design-based survey methods to causal and transportability questions, validated against NHANES. Methods paper in preparation.
Publications
Peer-reviewed work in observational and survey methodology, small area estimation, population health, and machine learning.
Get in touch
Open to senior roles in survey methodology, small-area estimation, and population-health measurement — at statistics offices, health agencies, foundations, research institutes, and organizations building serious data platforms. Full-time or contract. Also open to real-world-evidence work where design-based methods and transportability matter.
Also consulting on survey design, small area estimation, and statistical capacity for international agencies and research institutions. Bilingual English/French. I respond to all inquiries within two business days.
Download CV (PDF)I'll respond within two business days.