Software Factory Kit
My open-source operating model for AI-native delivery: phase-specific skills, human decision gates, automated quality checks, and a feedback loop for failures.
PROJECT ARCHIVE 004 / DEEPER DIVE
A closer look at the products, systems, and experiments behind my work. Browse by kind, watch a real walkthrough where available, and open the source.
Showing 13 projects
My open-source operating model for AI-native delivery: phase-specific skills, human decision gates, automated quality checks, and a feedback loop for failures.
A cross-venue prediction-market matching and order-routing simulation. Built in Go for a PEAK6 Labs partner project.
An in-EHR store slice where provider orders, patient payments, fees, and payouts reconcile in an auditable ledger.
A Solana market prototype with an on-chain order book, Pyth settlement, and a Next.js interface.
A driveway marketplace for crowded cities, with Android screens and Google Maps, Places, and Firebase flows.
A hackathon project that translates job-application emails into a simple status dashboard.
A full-stack task manager for learning product development from authentication through deployment.
A restaurant-search web app with an Angular front end, .NET backend, and PostgreSQL database.
A 2017 “I Stand With Immigrants” hackathon project centered on community storytelling.
A University at Buffalo hackathon experiment connecting a voice assistant with campus information.
An embedded experiment that translates sound into a responsive lighting setup.
Scripts for organizing current CAD documents while retaining approved sets and bulletin history.
A classroom tool for selecting students at random, built as an early client project.
Try a different search or category.
Looking for experiments and source code beyond this selection?
Browse every GitHub repository ↗HOW I BUILD 005 / AI-NATIVE WORKFLOW
Software Factory Kit gives AI-assisted work a repeatable structure. I define the intent and make decisions at gates; agents handle bounded investigation, implementation, and tests; independent checks verify the result.
Sharpen the brief, gather evidence, and write requirements before implementation.
sharpen-brief → write-prdMap the architecture, interfaces, acceptance criteria, and work slices.
design-hld → slice-tasksAssign work to agents with explicit file ownership and a clear definition of done.
implement-taskRun quality and security checks, review the diff, and record failures to improve the kit.
verifier → route-failureApproval gates cover product intent, architecture, milestone decisions, and shipping. The repository documents the gates, skills, scripts, and the evolving failure log.
Read how I build ↗