Engineering expertise
AI-Augmented Software Engineering
I use AI as an engineering assistant across the software lifecycle, from understanding an existing codebase to reviewing and validating a change. AI can help explore options, draft implementation, investigate failures, and suggest tests. I make the technical decisions, inspect the resulting code, and verify that it behaves correctly before it is delivered.
Tools I work with
- Claude
- Cursor
- OpenAI Codex
- GitHub Copilot
- LangGraph
- MCP
- TypeScript
- Python
My AI-assisted engineering workflow
AI helps me investigate and iterate. I retain ownership of architecture, implementation decisions, code review, and validation.
STEP 01
Understand
Clarify requirements, inspect the relevant code, trace dependencies, and establish the expected behavior.
STEP 02
Plan
Use AI to explore implementation options and risks, then choose an approach that fits the existing architecture.
STEP 03
Implement
Use coding assistants for focused implementation and refactoring while keeping changes scoped and understandable.
STEP 04
Debug
Use AI to help trace errors and identify likely causes, then verify the diagnosis against code and runtime behavior.
STEP 05
Review and test
Inspect generated changes, check edge cases, and run relevant tests to confirm expected behavior.
STEP 06
Deliver
Refine the implementation and review the final changes for maintainability, security, and production readiness.
Related project work
CRM Platform
An all-in-one CRM platform connecting customer records, sales pipelines, proposals, communication, scheduling, marketing, automation, and AI assistance in one workspace.
Read the case studySheet Parser
An AI-assisted workflow that interprets inconsistent vendor spreadsheets, validates proposed pricing, and routes uncertain values for review before they reach shared catalogs or project estimates.
Read the case studyAV Design & Engineering Suite
A commercial AV design application for creating floor plans, signal-flow diagrams, and rack layouts. Shared equipment data connects the views and supports project documentation.
Read the case studyRelated technical notes
AI / Agentic
Testing AI-Generated Code with an Independent Oracle
Verify generated changes with behavior-based assertions, failure cases, boundary tests, and deliberate checks that the tests can fail.
AI / ML
Human Review for AI Data: Design the Acceptance Boundary
Make extracted data reviewable with source evidence, field-level corrections, revision checks, and an explicit transition into accepted records.
Architecture
Start a SaaS Backend with Modules Before Adding Microservices
Use explicit ownership and small module interfaces first. Split deployment when an observed constraint justifies the network boundary.