Product-minded software builder
·5 wk▂▄█▃▂▃▂▁▁▁▁▃▁▁▁▁▂▁▁▂▂▁▁▁▂▁▂▁▁▁▁▁▁▁▁AI tools, educational apps, community products, and self-hosted systems — built with product clarity, design taste, automation, and full-stack engineering.
Useful
A free-by-default audit toolkit that emits deterministic JSON and lets the model write the prose.
See it inSEO/GEO Agent
Reliable
Integer-cents settlements, RLS predicates proven by probe, and a deletion state machine you can sleep through.
See it inSetOff
Private
Per-member opt-in baked into the schema; sharing and exports filter through visibility flags.
See it inKinGrove
Impactful
Built for neurodivergent learners who'd be left behind by generic edtech.
See it inFractionFriends
Featured Work
A curated set with enough polish, momentum, and product clarity to represent my current work.
A gentle full-stack fraction-reasoning app with spaced review, admin tooling, and child-centered UX for neurodivergent learners.
Combines education, product design, full-stack engineering, and AI-assisted learning into a cohesive app.
A privacy-conscious family network with opt-in profile sharing, self-hosted web deployment, and Supabase/Postgres infrastructure.
Demonstrates product strategy, privacy thinking, brand polish, deployment discipline, and operational maturity.
A genre-native romantasy book companion for the BookTok crowd — a polished brand, a live landing page that showcases the real product (rendered book spines and in-app mobile screens, not stock art), and a waitlist backend, now extended with a v0.1.0 mobile app prototype: a spine-out wooden bookshelf with a 12-template spine renderer, seven switchable shelf themes that re-skin every book, and a procedural-free / AI-baked-premium 'special editions' monetization model.
Develops a product end-to-end — from audience insight through brand identity and backend to a themeable mobile UI with a built-in free/premium model.
A unified trip dashboard for friend groups and families — itinerary building, expense splitting, and group decisions in one product, rather than four single-purpose tools and a group chat. Carries Skipper, an in-app AI co-planner that drafts itinerary items and polls for the crew to approve, and now a complete crew lifecycle: a member can leave and come back, an organizer can hand the trip over, and a guest who signs up keeps everything they already planned.
Combines product strategy, dense data modeling (RLS, settlement algorithm, multi-phase deletion flow), and operational lessons learned the hard way — including a privilege-escalation hole found on production and closed by migration — into one full-stack project.
A technical auditing engine for search and generative-engine visibility, live as a public self-serve web app at audit.corycopeland.dev — free scoring, accounts, delta tracking, and a tested Stripe billing layer staged for cutover — plus its own CLI-driven lead-generation and outreach engine. Every scoring rule now has to clear a release-gate corpus in CI before it counts as shipped.
Shows practical automation, crawler design, SSRF-safe architecture — including a pinning forward proxy that keeps the headless-Chromium render path from being redirected onto internal hosts — provider-abstracted billing, and applied AI-era content strategy, from a Claude Code skill to a deployed product.
An independent AI systems studio for small businesses and founders — lead capture, follow-up, and internal dashboards built as focused workflows, not a platform. The brand site now funnels to a dedicated AI Front Desk offer page, backed by an internal CRM with touch logging and a per-deal activity timeline, a quote/contract engine, and a demo pipeline that has produced named, click-through front-desk demos with video for more than a dozen real prospects.
Combines product positioning, a productized offer page, an internal ops app with a pricing/contract/delivery pipeline, and an automated personalized-demo system into one operating business.
An educational personal-finance organizer — explicitly never an advisor. Organize income, expenses, debts, and goals, see household-finance metrics, and generate a guardrailed AI-review prompt. Now local-first and encrypted at rest behind a passphrase, with a seeded Monte Carlo forecast that reports an uncertainty band instead of one confident line, five switchable skins, and its own prerendered marketing site.
Takes a privacy-first claim seriously enough to implement it: AES-GCM at rest with a runtime-calibrated PBKDF2 derivation, a boot state machine covering plaintext migration and corrupt-store recovery, and forecasting that refuses to overstate its own certainty.
Working in the open
Most of these projects ship from the same desk and the same small self-hosted stack. The constraints are the point.
Featured case study
A gentle full-stack fraction-reasoning app with spaced review, admin tooling, and child-centered UX for neurodivergent learners.
Labs
Prototypes and edges of what I'm exploring next. Visible, but secondary.
A calmer supplement tracking app for people who take this seriously. Smart scheduling, interaction-aware, science-backed — built for the over-40 stack.
A mobile-first AI lyric video tool concept under BeaconBridge Labs.
About
I'm Cory Copeland — a builder who cares about product clarity, humane interfaces, useful automation, and systems that can actually be operated after the fun prototype phase ends.
More about CoryWriting
Scattered on Purpose is my Substack for reflection. Build Notes will be the technical companion: architecture decisions, self-hosting, and AI-assisted development.
