ZekeLabs

Software products, architected and built with AI.

An independent studio that takes software from spec to running product. AI does the building, every change is checked, and nothing ships until it is approved.

See the workHow it's built

The Digital Novice: each daily issue is written, illustrated, checked and published with no one at the keyboard.

Selected work

  1. A published issue of The Digital Novice: the headline AI-Generated Voices Are Now Placing Fake Emergency Calls to 911 Dispatchers, dated November 11, 2026, above an AI-drawn header image of the raccoon mascot next to a dispatch alarm.

    Live

    The Digital Novice

    A fully autonomous AI newsletter. Every day it reads trusted sources, writes and illustrates the issue, and publishes only if a second model passes it.

    Architect and operator. Designed the engine, its gates and its specs; AI wrote all of the code.

    Next.js, Claude API, Supabase, Resend, Stripe, Replicate, Vercel

    Read the full project

  2. The Kalsys connect page: the headline Bring your memory to your AI client, a short description of how the assistant proposes, Kalsys checks and writes, and the owner approves every save, and a notice that the server is in a private build.

    In build

    Kalsys

    One memory shared across AI assistants. The assistant proposes a change, Kalsys checks it against written rules, and nothing is saved until the owner approves.

    Architect and operator. Specified who may write to memory and which actions an AI can take; directed Claude through the build and every review.

    TypeScript, MCP, Postgres, Supabase Auth, Railway

    Read the full project

  3. Customer portalNext.js appSupabase auth and dataBackground workers

    Private

    Utility billing software

    Billing software for a water utility with 200+ accounts, in build: recurring billing with autopay, a customer portal and back-office tools.

    Specified the system and directs its AI build.

    Next.js, TypeScript, Supabase, Cloudflare Workers

    Read the full project

Legacy game systems

Legacy game software, studied and repaired without source code: reading the binaries, tracing the network, recovering what was lost, and redrawing what had aged.

  1. Legacy client and serverAI agents trace and patchWritten spec per fixChecked against the running client

    In build

    Legacy online RPG restoration

    A 2000s online RPG restored on a private test setup without source code: the client analyzed and patched, lost text recovered, interface art redrawn with AI.

    Architect and director. Sets each investigation and fix; AI agents trace, patch and test; every change is checked against the running client.

    Python, x86 binary patching, Packet tracing, Database tracing, Image models

    Read the full project

One method on every project

  1. Specify

    ZekeLabs writes the spec and the rules the build has to follow.

  2. Build

    An AI model builds and tests against that spec.

  3. Check

    A second model that hasn't seen the work reviews it and marks what must change.

  4. Approve

    No change to a system is saved or shipped until ZekeLabs approves that exact change.

Talk about a role or a build

hello@zekelabs.dev