Caleb Lanting
Full-stack and AI developer

I build products and AI automation people actually use.

I'm Caleb Lanting, a full-stack and AI developer in Riverside, California. Claude Code and Cursor are at the center of how I work: agents write the code, and I decide what to build, review every change and make sure it works.

Selected work

Each project has its own page: what it is, who it's for, the hard part, and how I built it.

LedgR

A calm CRM for small financial-advisor firms.

Clients, pipeline, follow-ups, emails and texts in one place, at the desk and on the phone. Notes set their own reminders, read by rules on the device first and Claude Haiku second.

Live by invitationWeb, iPhone, Android
TrackR

A companion app for coaster and theme-park fans.

Log every ride, then plan a park day with live waits, hours and weather. It runs on a data platform that records where every fact came from and what its license allows.

Beta on TestFlightReact Native, Python, PostgreSQL
How I build

How I build with AI agents.

An always-on team of agents, and a bug pipeline where one agent reproduces, one fixes, and an independent one re-runs the steps before anything counts as fixed.

In daily useClaude Agent SDK
AI evaluation

Grading AI writing with evidence, not vibes.

A judge that must quote every flaw it charges for, five runs and a median, and a fixer that only touches what was quoted. Plus a blind test of whether more agents build better software.

PythonLLM judges
iAttend

Attendance for an after-school program at 12 schools.

Instructors take attendance on their phones, admins see every site live, and the district that funds the program gets its own portal. About 500 students.

Launched March 2026Next.js, Firebase
Lead responder

One tap from a new lead to a sent text.

For Pete's Holiday Lighting: every quote request gets a reply drafted from the owner's own scripts, and one tap opens Messages with it filled in.

LiveCloud Functions

More work

Smaller builds, early sites, and the game I'm making now.

How I work

Agents write the code. I write the spec, make the product calls, and review every change. Then something other than the agent that wrote it checks that it works, before anything ships.

When something slips through, I don't just fix the bug. I fix whatever let it through.

How I build with AI agents

Want to build something together?

I'm looking for a team to build with full time. Email is the fastest way to reach me.