SYSTEM OPERATIONAL gateway · 4 providers 1,284 req/min rejected 0.3% p95 420ms tokens 2.67M / 24h
AI Native Technology Company

Unify how the enterprise
统一connects, governs & runs AI

AI products · open-source projects · professional engineering services

QZ Technology delivers productizable, self-hostable and continuously evolving capabilities across AI service governance, workflow automation, software engineering and enterprise digitalization.

AegisAI Service Governance
TrigixAI Workflow Automation
DeliveryEngineering Delivery
aegis · ai-gateway LIVE
请求 / min1,284
拒绝率0.3%
P95420ms
tokens/24h2.67M
model routing · live
audit log · live12,840
Product

Two core products, one engineering stack

Two core products, one engineering stack

Connect AI cleanly, govern it clearly

Aegis centrally manages API keys, providers, model routing, concurrency, token budgets and audit records for internal AI applications, agents and developer toolchains.

01
Unified AI Entry Point
Applications connect through an OpenAI-compatible API while upstream models and credentials stay centrally hosted.
02
Model Routing & Policies
Route requests by business, team, cost and model capability to reduce integration complexity.
03
Budget, Audit & Observability
Track call chains, token usage, denied requests and key configuration changes.
request governance flow
应用 / 智能体OpenAI-compatible
Aegis Gateway● active
密钥隔离路由策略Token 预算审计
Anthropic
OpenAI
阿里 / 智谱
Open Source · MIT

Trigix: an AI-native workflow automation platform

Trigix connects AI models, APIs, databases and SaaS tools through a visual drag-and-drop canvas to build, run and monitor complex business workflows. 180 node types, a Rust DAG execution engine (about 5–10x the throughput of Node.js) and native MCP integration distill our engineering in AI productization and enterprise-grade delivery.

180Node Types
RustDAG Execution Engine
MITLicense
AI 原生 · 8 内置大模型JWT + RBAC · 多租户Docker · K8s Helm
workflow.yaml
workflow:
  trigger: webhook.hmac_sha256
  canvas: react_flow
  executor: rust_dag_engine
  control: [parallel, foreach, retry]

nodes:
  total: 180
  ai: [Claude, GPT, Gemini, Qwen]
  integrations: [GitHub, Jira, Slack]

runtime:
  auth: JWT + RBAC
  store: postgres + pgvector
  deploy: docker + k8s_helm
AI Capability

A technology stack from AI products to business systems

A full stack — from AI products to business systems

01

AI Service Governance

Unified model entry, key isolation, provider management, budget control and call auditing.

02

Business System Engineering

Build web apps, admin consoles, API services and data pipelines around business processes.

03

Quality & Security Validation

Software testing, compliance support, acceptance testing, risk review and delivery materials.

04

Self-hosting & Operations

Server deployment, containerization, log monitoring, upgrades and ongoing support.

Engineering Flow

An engineering flow that delivers solutions to a running state

From proposal to a running system

01

Scope & Boundary

Define business goals, AI usage boundaries, data sources, permissions and delivery scope.

02

Product & Architecture

Decide between Aegis, Trigix, custom development or a mix, and produce deployment and API designs.

03

Build, Integrate & Validate

Complete development config, system integration, test fixes, permission setup and acceptance prep in phases.

04

Launch, Train & Evolve

Handle deployment cutover, runtime checks, user training, documentation and ongoing optimization.

Ready to Build

Planning AI access, workflow automation or a new business system?

Share your business scenario and current systems, and we'll help decide between productized access, custom development or phased delivery.