Freddy Ray Drennan — cv
Download PDF
Freddy Ray Drennan — Curriculum Vitae

Freddy Ray Drennan — Curriculum Vitae

Updated: September 2025

812 Cameron Dr, Severance CO 80550 • +1 254 931 8313 • drennanfreddy@gmail.com
LinkedIn: freddydrennan • GitLab: @fdrennan • GitHub: @fdrennan

Summary

Engineer focused on R platforms and high-performance computing for research & development. I build reliable data applications and maintain reproducible, validated environments across shared clusters and cloud. I bridge source code, CI/CD, “build from source”, and runtime operations (modules, schedulers, and internal publishing).

Core Competencies

  • Computational Platforms: environment modules, batch scheduling, multi-cluster site libraries, access controls
  • Build from Source: compile R and system components from scratch; resolve ABI/compiler/linker issues (e.g., glibc & toolchain mismatches)
  • Diagnostics & Reliability: root-cause analysis, standardized health checks, incident response & MTTR reduction
  • Applications & Reporting: production web UIs (R/Shiny) and governed report pipelines with review workflows
  • Ops & Governance: CI/CD, packaging, artifact promotion, validation/SOP mindset
  • Tooling: R (primary), Rust (platform tooling), shell; experienced with Python/SAS

Professional Experience

Senior Developer — HPC & Clinical Tooling · Roche / Genentech (CEDAR/ROCS) · Nov 2024–Present
- Operate data apps and reporting on shared HPC with environment modules and batch schedulers.
- Compile R stacks and dependent components from source; remove ABI/runtime drift across clusters.
- Maintain CI pipelines, documentation, and site-library promotion; add health checks and diagnostics to cut MTTR.
- Incident leadership: First responder to a breaking change in R curl; coordinated with maintainers to triage and land a fix.

Senior Application Developer · BioMarin Pharmaceutical · 2020–Nov 2024
- Delivered internal analysis apps and reporting pipelines for clinical/statistical work.
- Built integrity-monitoring and analysis utilities; containerized execution with CI for dependable runs.
- Migrated legacy analysis toward reproducible, cloud-friendly deployments (Posit Connect + standalone).

Founder — ndexr.io · 2021–Present
- Provision secure, standards-aligned stacks on a customer’s AWS (EC2, Route53, EIP) with automated domain setup, TLS (ACME/Let’s Encrypt), NGINX routing, and Docker Compose—no vendor lock-in.
- Support single-server, multi-domain deployments with zero-downtime restarts, health probes, and one-click rollback.
- Ship a lean web console for key pairs, security groups, DNS, logs, and per-app downloads; guard-railed SSH helpers and copy-to-clipboard ops.
- Deliver app templates (R/Shiny, Quarto sites, static assets) with CI hooks for build/publish, versioned configs, and reproducible environments.
- Add observability: access/health dashboards, geo usage summaries (GeoLite2/IP2Location), and basic cost views.
- Harden by default: dual-stack IPv4/IPv6, strict CSP, OAuth2 login (Google), and sane NGINX defaults.

Business Analytics Manager · LogistiCare · 2019–2020
- Containerized decision service for rider–provider matching; addressed class imbalance at scale.
- Deployed EC2-hosted R APIs to improve assignment accuracy and reduce cost.

Data Scientist / Data Optimization Engineer (Adtaxi / Digital First Media) · 2017–2020
- Built and operated Quantum LP: a linear-programming engine (R, lpSolve) that maximizes expected conversions subject to lifetime budget, channel minima, adgroup floors/ceilings, and legacy spend.
- Shipped a trader-facing Shiny console (DT/plotly/flexdashboard) around the solver with CSV export, infeasibility diagnostics, and guardrails for pacing and campaign constraints.
- Integrated with internal MongoDB collections and campaign IDs; supported large campaigns (thousands of adgroups) with fast solves and back-testing workflows.
- Delivered automated reallocation across Facebook / Search / Display, improving VDP/Cost (ROAS) while honoring governance rules.
- Administered the internal R platform (Connect/Server/Package Manager) and CI for LP + reporting code; published quant-oriented helper packages and tests.

Earlier Clinical Experience (condensed)
- Trauma Nurse (Afghanistan) — 555th FST, U.S. Army
- Med/Surg Nurse — Darnall Army Medical Center
- Pediatric & Foster-Care Nursing — Epic Healthcare; A World For Children

Selected Projects & Systems (Platform Focus)

  • cedrus (Rust CLI): environment/module discovery, multi-repo ops, onboarding, and diagnostics for shared HPC.
  • Site Libraries Orchestration: source-built R + system components, promoted across clusters via CI with deterministic results.
  • HPC App Pattern: thin-server R apps with externalized rendering and stable publishing on internal platforms.
  • Governed Environments: standardized env-vars, health checks, and “glimpses” to surface failures early in CI.

Impact Highlights

  • Delivered deterministic R environments across multiple clusters with controlled promotion paths.
  • Resolved glibc/compiler/toolchain mismatches without brittle pinning.
  • Led incident response for a critical upstream break in R curl; coordinated maintainer-level fix.
  • Reduced onboarding time by codifying environment discovery and diagnostics into CLI utilities.
  • Standardized CI and environment checks for modules, applications, and documentation artifacts.

Clinical & Standards

Working knowledge of end-to-end clinical data flows and ADaM-aligned analysis datasets; audit-ready reporting practices and review pipelines.

Education

  • The University of Texas at Austin — BBA Quantitative Finance, Minor in Mathematics
  • Army Medical Department Center & Schools — LPN (Compact License); EMT (Combat Medic)

Certifications & Military Awards

  • Combat Action Badge
  • Army Trauma Training Course — Ryder Trauma Center (scenario-driven clinical exercises)

Keywords

HPC • environment modules • batch schedulers • site libraries • reproducibility • CI/CD • diagnostics • incident response • build from source • cloud deployments • validation/SOP • internal publishing • platform tooling (Rust) • data applications (R)