Multi-Agent AI Creation Infrastructure

AI-First Platform for
Video, Apps & Agents

Q-Star AI is building a multi-agent AI creation infrastructure for video production, app generation, interactive games, autonomous agent workflows, and automated media โ€” unified under one intelligent platform.

4
AI Products Live
15+
AI Models Integrated
>50%
AI/ML Workload Share
Est. 2023
Guangzhou, China
Google Cloud Startup Credits Use Case

Preparing to Consolidate on Google Cloud

Q-Star AI already operates AI generation workflows across video production, app generation, autonomous agents, and automated media production. We are preparing to consolidate these workloads on Google Cloud, with Vertex AI / Gemini as the core AI layer.

๐Ÿง 

Vertex AI / Gemini API

Planning, routing, code, text, and multimodal generation across all four products.

๐Ÿš€

Cloud Run

API services, agent workers, webhooks, and lightweight orchestration at scale.

โš™๏ธ

GKE

Container orchestration for high-volume generation workers and long-running services.

๐Ÿ“ฆ

Cloud Storage

Generated videos, images, APKs, prompt archives, and workflow artifacts.

๐Ÿ“จ

Pub/Sub

Async task dispatch, rendering queues, retries, and workload fan-out.

๐Ÿ“Š

BigQuery + Monitoring

Usage analytics, workload forecasting, cost reporting, and reliability dashboards.

Full Google Cloud Use Case โ†’

One Platform. Five AI Creation Surfaces.

We don't build single-purpose AI tools. We build the infrastructure layer that makes AI creation repeatable, scalable, and programmable.

Video Production

Drama Engine

Full-stack AI content pipeline for short-form video. From prompt to published video โ€” automatically.

  • Brief โ†’ script โ†’ TTS โ†’ render โ†’ delivery
  • 6 voice profiles, male & female
  • AI-generated BGM from licensed library
  • Subtitle auto-burn with timing sync
  • MP4 output delivered to CDN
Learn more โ†’
Autonomous Agents

AgentOS

Multi-agent orchestration that decomposes tasks, delegates to specialized sub-agents, and executes with real tools.

  • Hierarchical planning & task decomposition
  • Web search, code exec, file ops, API calls
  • Persistent memory across sessions
  • Hermes automation sub-agent
  • Telegram human-in-the-loop interface
Learn more โ†’
App Generation

AppForge

Describe your app in natural language. Get a production-ready iOS + Android app. No coding required.

  • iOS & Android generation from text
  • AI-generated screenshots & assets
  • Screenshot-to-code conversion
  • APK packaging & delivery pipeline
Learn more โ†’
Consumer AI

ClipAI + AI Music

Mobile-first AI short video app and AI music composition โ€” consumer-facing layer on top of our infrastructure.

  • ClipAI: AI short video (iOS + Android)
  • AI Music: WeChat Mini-Program, live
  • Powered by Drama Engine + Model Router
Learn more โ†’
Interactive Games

AI Game Engine

Config-driven interactive simulation engine for AI-generated 360ยฐ panorama escape rooms and adventure games.

  • 360ยฐ panorama scene generation
  • Hotspots, inventory, and puzzle logic
  • AI-generated panoramas, icons, and audio
  • Castle of Last Clues โ€” software copyright submitted
  • H5 browser-native, no app install required
Play Demo โ†’

How It Works

Every product runs on the same multi-agent infrastructure. Here's what happens under the hood when a user creates with Q-Star AI.

1

User Intent

Text prompt, voice command, or UI interaction from any client surface โ€” mobile app, web, or API.

2

AgentOS Orchestration

Decomposes intent into an execution plan. Assigns sub-tasks to specialized agents with the right tools.

3

Intelligent Model Routing

Routes to Vertex AI / Gemini, or external models (Claude, GPT, MiMo) with load balancing and automatic failover.

4

Processing Pipeline

TTS synthesis, video rendering, image generation, audio mixing โ€” all automated and parallelized.

5

Cloud Delivery

Output delivered via CDN. Video files, app packages, or API responses โ€” ready instantly.

AI-First Workload Profile

AI/ML Workloads >50% of Cloud Usage

AI/ML workloads are expected to represent more than 50% of Q-Star AI's cloud usage, driven by multi-agent planning, model routing, multimodal generation, video rendering, app generation, TTS, and automated media production.

WorkloadAI/ML RelevanceGoogle Cloud Target
Agent planningLLM task decomposition and routingVertex AI / Gemini
Video generationScript, voice, render, subtitles, output assemblyCloud Run, GKE, GPU, Cloud Storage
App generationUI/code generation, build packaging, assetsGemini, Cloud Run, Cloud Storage
Model routingMulti-model orchestration and failoverVertex AI, Gemini, Monitoring
Media processingTTS, image/video/audio generationGPU workers, GKE
AnalyticsUsage, costs, workload forecastingBigQuery
Full Workload Forecast & Migration Plan โ†’

Real Products. Real Infrastructure.

We ship. Every milestone below is live in production.

MilestoneStatus
AI Music Mini-Program live on WeChatโ— Live
ClipAI submitted to App Storeโ— In Review
Drama Engine generating production videosโ— Live
AgentOS running daily autonomous workflowsโ— Live
iOS + Android App Generation capabilityโ— Operational
Multi-model intelligent gateway (15+ models)โ— Production
ๅนฟๅทžๅฅ‡่ฐท็ง‘ๆŠ€ๆœ‰้™ๅ…ฌๅธ incorporated2023

Infrastructure: Current State โ†’ Google Cloud Target

Q-Star AI currently operates on self-managed cloud infrastructure. We are preparing to migrate and scale these workloads on Google Cloud.

Current Operating Infrastructure

๐Ÿ–ฅ๏ธ

4 Vultr Servers

Production workloads, distributed across regions.

๐Ÿณ

Docker Compose

Containerized workloads, designed for GKE migration.

๐Ÿ“ฆ

Cloudflare R2

Generated videos, APKs, images, audio assets.

๐ŸŒ

Qiniu CDN

China-optimized content delivery.

Google Cloud Migration Target

๐Ÿง 

Vertex AI / Gemini

Core AI layer for planning, routing, generation.

๐Ÿš€

Cloud Run / GKE

Scalable agent workers and generation services.

๐Ÿ’พ

Cloud Storage

Generated media, APKs, prompt archives.

๐Ÿ“จ

Pub/Sub + BigQuery

Async queues, analytics, cost forecasting.

Let's Build the Future of AI Creation

Investor and startup ecosystem partner discussions are in progress. We're open to partnerships and conversations.

Contact Us Google Cloud Use Case Evidence โ†’