Moonray Studio
Available for new projects  ·  Kuala Lumpur, working worldwide
AI systems  /  automation  /  products

Systems that ship in weeks, not quarters.

Moonray Studio builds AI systems and products for companies that want to move fast. One builder with AI leverage, doing architecture, build, and handover, at a fraction of agency pricing.

Tell me what you are building Three AI products designed and shipped this year

Three ways to work together

Fixed scope, fixed price, agreed before anything starts. No hourly meters, no open ended discovery phase.

Automation

Automation builds

Chatbots for Telegram and WhatsApp, content pipelines, lead handling, document processing, internal tools.

From $3,0002 to 4 weeks
Products

Custom AI products

Conversational products, multi step AI engines, retrieval systems, payment and database integrations.

From $5,0003 to 4 weeks
MVP

MVP builds

For non technical founders. Your idea to a working product people can use, not a prototype that stalls.

From $8,0003 to 5 weeks
Every build ships with ongoing support included. You get a working system, not a slide deck.

Built, shipped, and running

Three AI products designed and built solo this year. Architecture, code, and deployment, all of it.

Live in production

AI Memoir Platform

An AI companion holds voice conversations in Persian over months, remembers everything told, works out where one story ends and the next begins, and assembles it into a printed hardcover book. The polish engine keeps the speaker's exact phrasing and is never permitted to invent a fact.

Python · FastAPI · Claude API · pgvector · Supabase · Next.js · Cloudflare Workers · Stripe · Print API
Per family encryption · Hash chained audit log · Full data deletion across every provider
8,500 lines

AI Production Platform

Pre production and production in one system, carrying a project from story through script, image, and video without drifting between stages. The hard part is consistency: locked reference artifacts with stable IDs, gate validation between phases, and inheritance contracts so a later step cannot silently redefine an earlier decision.

React · Claude API · IndexedDB · Model agnostic generation layer
Two second loop

Real Time AI Story Engine

A child speaks an idea and watches it become a scene in about two seconds, then decides what happens next. Image generation and story state updates run in parallel while narration plays, so the wait disappears rather than being shortened.

Next.js · Claude API · SDXL Turbo · Web Speech API · Vercel KV · Stripe

Why one person is faster

An agency bills you for an account manager, a project manager, and a team of four, then spends the first two weeks introducing them to each other.

I do the architecture, the build, and the handover myself, with AI leverage covering the work a junior team used to do. Nothing is lost in translation because there is no translation.

That is why delivery is measured in weeks and why the price is what it is.

Tell me what you are building.

I will tell you honestly whether I can help, what it will cost, and how long it will take. If it is not a fit, I will say so.

Amir Asl, founder
Moonray PLT · Kuala Lumpur, Malaysia
Working with clients worldwide, priced in USD