Statistical Computing & Scientific Software

R (Primary Engineering Language)
Advanced package development and scientific computing infrastructure.

  • Production-grade CRAN package authoring and maintenance
  • S4 and R6 object-oriented design
  • Engine wrappers and cross-language interfaces (Rcpp / C++)
  • Shiny-based interactive modeling systems
  • Simulation workflows and reproducible statistical pipelines

Python
Algorithmic model search and automation pipelines.

  • pyDarwin optimization framework
  • ML-driven model selection integration
  • Cross-engine orchestration (NONMEM / NLME)

Core competencies include regression modeling, simulation-based inference, and structured statistical workflows.

Modeling Engines & Domain Infrastructure

Deep integration experience across enterprise pharmacometric systems:

  • Phoenix NLME — estimation, simulation, covariate modeling, engine-level workflow integration
  • NONMEM — population PK/PD modeling and ML-driven search integration
  • Simcyp Simulator — PBPK/PD simulation automation via R-based orchestration
  • PML (Pharmacometric Modeling Language) — specification, generation, and configuration pipelines

Architectural focus:
Engine abstraction layers, cross-engine workflows, API wrappers, and reproducible modeling infrastructure.

Optimization & Simulation Systems

Algorithmic and performance-oriented modeling infrastructure:

  • Automatic differentiation for NLME estimation (ADPO)
  • FOCE, Laplacian, and AGQ estimation workflows
  • Objective function design (OFV / -2LL minimization)
  • Multi-objective genetic algorithms (MOGA)
  • Gaussian processes, random forests, GBRT, and PSO optimization
  • Large-scale model search validation (1.57M model space benchmarks)
  • Monte Carlo simulation pipelines and VPC automation
  • Parallel and grid-based execution orchestration

Focus: translating mathematical optimization into production-grade estimation systems.

Cloud-Native Modeling Architecture

Enterprise modeling platforms delivered via modern full-stack systems.

Backend & Services:

  • Node/Express and Java-based service architectures
  • REST API design and engine integration layers
  • OAuth2 / OIDC-aware identity and licensing integration
  • Containerized execution environments (Docker)
  • CI/CD-enabled deployment workflows
  • AWS-based deployment contexts

Emphasis on scalable modeling workflow delivery rather than generic web application development.

Front-End & Interactive Systems

User-facing modeling interfaces built for scientific workflows.

  • React / TypeScript component architecture
  • Modeling workflow state management
  • UX design for configuration-heavy modeling systems
  • R Shiny production-grade interactive applications
  • Diagnostic visualization systems (ggplot2 ecosystem)

Bridging modeling engines with usable, extensible interface layers.

Infrastructure & Workflow Engineering

Operational systems supporting modeling platforms:

  • CI/CD pipelines (GitHub Actions, Jenkins)
  • Version-controlled, multi-repository workflows
  • Grid workload managers and distributed execution systems
  • Containerized build and deployment strategies
  • Cross-platform modeling execution environments

Focus: reproducibility, scalability, and system-level robustness.