Selected work areas

Skills, methods, and engineering context.

A capability-based view of my work across AI, data, and engineering. Employer-sensitive details, internal systems, and proprietary data are intentionally excluded.

01 · Engineering context

Domain knowledge changes the work.

Applied and exploratory work where engineering meaning, data quality, validation, and technical judgment matter as much as the model.

Current professional context

Generalized public summary

Aerospace AI & Airframe Engineering Workflows

Applied AI systems at a major aircraft manufacturer, including fuselage, airframe, and major aircraft component engineering contexts.

  • Aerospace AI
  • Applied AI systems
  • Fuselage & airframe workflows
  • Technical collaboration

Collaboration includes engineers and technical stakeholders working with major aircraft components. No internal project names, systems, datasets, aircraft details, or engineering results are represented here.

Scientific computing

Personal proof of concept

Scientific ML & Surrogate Modeling

Exploring how experimental design, simulation-oriented data, and surrogate models can support faster analysis while preserving clear validation boundaries.

  • Python
  • Scientific ML
  • Design of experiments
  • Surrogate modeling

NasgroPy is a small personal Python proof of concept used to explore these methods. It uses no employer code, proprietary data, licensed outputs, or restricted engineering results.

02 · Reliable AI systems

Evidence and expert judgment belong in the workflow.

Evaluation engineering practices for understanding model behavior, retrieval quality, failure modes, and the role of human review.

Evaluation engineering

Capability area

AI Evaluation & Retrieval Workflows

Designing repeatable ways to test LLM behavior, inspect retrieved context, analyze failures, and connect evidence to practical operating decisions.

  • LLM evaluation
  • Retrieval quality
  • Failure analysis
  • Trace review

Review & feedback

Capability area

Human-in-the-Loop AI Systems

Structuring annotation, expert review, and feedback loops so uncertainty remains visible and decision ownership stays with the right people.

  • Expert review
  • Annotation
  • Feedback loops
  • Decision support

03 · Domain systems

Useful automation respects the source material.

Work around technical language, specialized documents, publishing, and repeatable workflows where generic automation is not enough.

Language & knowledge

Capability area

Domain NLP & Technical Communication

Turning specialized documents and terminology into structured, understandable information without hiding source provenance or the need for domain interpretation.

  • Domain NLP
  • Technical documents
  • Provenance
  • Structured communication

Independent infrastructure

Capability area

Automation & Publishing Systems

Building small, durable workflows for content preparation, review, static-site generation, and controlled delivery with low operational overhead.

  • Workflow automation
  • Static publishing
  • GitHub Actions
  • AWS delivery

Professional context

Interested in how I approach technical problems?

I am open to thoughtful professional conversations, future opportunities, and selected collaborations.