Codelucent
AI Design Studio · Kansas City & Tokyo

Ship the AI product your competitors keep demoing.

Codelucent is an AI design studio in Kansas City and Tokyo. Napkin to App Store, prompt to production: mobile apps, LLM systems, payments, cloud. We run our own apps, so we know exactly what breaks at 2 a.m. and how to build so it doesn't.

Shipped, not slide-deckedEvals on every changeOur own apps in the stores
Tori: a marketplace where travelers bring the world home
Akiya Listings: map search for vacant homes across Japan

Teams and products we have built with

Casca
Cyphr
Leasey
Curb Hero
Thirty Capital
Startland News
Particle Space
Google
Enigma
Boddle Learning
Bucs
Casca
Cyphr
Leasey
Curb Hero
Thirty Capital
Startland News
Particle Space
Google
Enigma
Boddle Learning
Bucs
Our apps

We eat our own cooking.

Seven products we design, ship, and run ourselves. Marketplaces, fintech rails, proptech, health, and vertical SaaS, live on iOS, Android, and the web.

See all apps
Scuber icon
Apps coming soon
Services marketplace

Scuber

Lost it in the water? Get it back.

iOS and Android apps launching soon

iOSAndroidWeb
Paystack icon
Live
Fintech · Infrastructure

Paystack

Payment rails built for agents and marketplaces.

WebAPI
Leasey icon
Live
Proptech · AI copilot

Leasey

Property management, run by chat.

Web
Results

Numbers, not adjectives.

4.2×
retrieval hit rate lift
>95%
extraction accuracy
~40%
support ticket deflection
Zero
cross-tenant data leakage
How it fits together

A production RAG pipeline.

Click a stage to see the code behind it.

Connect to your docs, tickets, warehouse, and databases. Versioned ingestion pipelines re-embed only what changed.

ingest.ts
// Versioned ingestion with source-version tracking
export async function ingest(source: Source) {
  const docs = await source.fetchSinceLastRun()

  for (const doc of docs) {
    const chunks = chunk(doc, {
      strategy: source.strategy,   // semantic | fixed | layout
      size: source.chunkSize,
    })

    await embed(chunks, {
      model: "voyage-3",
      cache: true,
      version: source.embeddingVersion,
    })
  }
}
Evaluations

Already have AI in production? Let's find out if it's lying to you.

We audit retrieval quality, faithfulness, safety, cost, latency, and drift on your real traffic, fix the worst offenders, and wire evals into CI so the fixes stick.

Explore evaluations
>95%
field-level extraction accuracy
3-5x
retrieval hit rate lift
<5%
hallucination rate target
30-60%
inference cost reduction
Process

Four steps. Zero mystery.

01

Discovery

We map your users, data, workflows, and constraints, and agree on what "good" looks like in measurable terms.

02

Design & architecture

Flows and interface, retrieval shape, model routing, infra, guardrails, and the eval plan that proves it.

03

Build & evaluate

Weekly demos. Every iteration gated by evals on your data, not vibes or vendor demos.

04

Ship & harden

Store launch or production deploy in your cloud, observability and cost controls wired, runbooks handed off.

Leasey Operations Team

“We were able to scale our tech operations with a fraction of the cost and time it would have taken to build in-house.”

Leasey Operations Team

Got an AI idea? Bring it. We'll tell you the truth about it.

30 minutes with a founder: what to build, what to skip, and what it should cost. No pitch deck, no account manager.

Other ways to reach us