jameschang.co [experience] [/now] résumé.pdf ↓

james chang / projects

// selected project artifacts

Real pages from real apps.

Static snapshots of internal tooling from three of my current products — roadmaps, changelogs, technical overviews, and how-it-works docs pulled directly from the running apps and mirrored here for anyone who wants to see the work without a login.

/01

aleph

Compliance SaaS for US importers — FDA FSVP, CPSIA/CPC, California Prop 65, and multi-state PFAS. 12 releases in 14 days. alephco.io ↗

Four-step process from SKU import to compliant output, plus a module-by-module comparison of what each regulation covers.

All four compliance modules with workflows, feature status, and six expansion verticals (apparel, cosmetics, furniture, electronics, supplements, automotive).

12 releases documenting the FSVP Pro → Aleph rebrand and the build-out from single-module to multi-cohort platform.

How one prompt became an executive compliance dashboard — the iterative process, the final spec, and the design decisions behind every metric.

/02

the fantastic leagues

Fantasy baseball platform with league-context AI. 26 feature modules, 730 passing tests, 84,400 lines of code. thefantasticleagues.com ↗

Eight league-context-aware AI features with Gemini primary + Claude fallback, plus a shared SQLite insights cache keyed on team & data hash.

Architecture, full stack, 26 feature modules, 49 database models, 62 build sessions over six months.

PostHog tracking, development velocity across 33 sessions, four key product questions.

Recent releases through v0.57 — admin dashboard, session tracking, error correlation, parallel-agent build workflow.

Two prompts became one — how an enterprise-scale spec evolved into a context-aware dashboard for a 14-user fantasy app.

/03

the judge tool

Digital KCBS BBQ competition judging. 16,424 lines, 113 tests, shipped in 20 days with transparent AI-assisted workflow documentation. thejudgetool.com ↗

Full build teardown: 16,424 LOC, 12 database models, 62 server actions, 113 tests, architecture diagrams, AI workflow.

8/10 health scorecard across 6 dimensions, 6-item risk register with mitigations, phase-by-phase roadmap.

14 releases in 20 days — full app built in one session, then security hardening, WCAG pass, and Vercel migration.

// snapshot — these are point-in-time mirrors. The live apps keep evolving — versions, counts, and roadmaps shift weekly. For the current state, follow the outbound links on each page.