Now

What I'm focused on, right now.

A snapshot of where my attention is. Updated periodically.

Last updated: May 2026 · Inspired by nownownow.com

Career

  • Exploring senior RWE / pharmacoepidemiology / HEOR roles at pharma, RWE platform vendors, and CROs — particularly those that combine observational methodology with platform engineering.
  • Top targets: Aetion, Flatiron Health, TriNetX, Tempus (RWE platform companies); Genmab, J&J, Pfizer, Novartis (pharma); IQVIA, Evidera, ICON (CROs).

Methods & research

  • Building svy-causal — the survey-weighted causal inference module of the svy ecosystem. IPTW, stabilized weights, doubly robust estimation, validated against NHANES.
  • Poster accepted at ISPE 2026 (Milan, August 29 – September 2): "Design-Based Causal Inference for Pharmacoepidemiology: Survey-Weighted IPTW and Doubly Robust Estimation in Python With NHANES Validation."
  • Studying ICH E9(R1), the FDA RWE framework guidance, and ISPE Good Pharmacoepidemiology Practices.

Platform

  • svyLab is at 327 passing tests and approaching public preview. Three hardening phases done; the lifecycle/classification authority model is settled. Currently finishing the dataset detail page redesign and the demo seed data for Ministry-of-Health and UN-agency demo orgs.
  • Frontend in Astro + Svelte 5; backend in Python (Litestar) + PostgreSQL + DuckDB + Redis; AI layer via svy-agents.

Consulting

  • Selectively. Recently completed: NBS Tanzania capacity assessment for the World Bank / SADC. INSBU Burundi small area estimation training (2 weeks). FeeLoST DHIS2-integrated platform for Ethiopia's Ministry of Health.
  • Open to short-term technical assistance on survey design, small area estimation, or RWE infrastructure questions. English or French.

Reading & thinking about

  • Target trial emulation in observational RWD; marginal structural models; competing risks under complex sampling.
  • How RWE platforms (Aetion, Flatiron, etc.) handle reproducibility, audit trails, and AI-assisted analysis. Where svyLab's design overlaps and where it differs.
  • The slow but accelerating convergence between survey-design-based inference and pharmacoepidemiology methods.

If something here resonates with what you're working on, get in touch .