Engineering the Future

Engineering partner for
AI and software that ships

Yejua is an AI agent and software group led by a full-stack engineer and CTO with more than twenty years in the field.

Architecture & Scalability

The structure running today is the one that scales

Design, deployment, operations and incident response run through one team. Decisions land quickly, and what gets built actually runs.

A architecture built to grow

Starting from a single-server environment, independent AI agents and complex automation pipelines — 130+ n8n workflows — were integrated into one coherent system, designed to expand into a multi-server cluster as traffic demands.

Unified LLM control · AI Gateway

We built our own high-performance gateway that governs multiple providers behind a single endpoint, handling call logging, usage ceilings and settlement in one place.

OpenAIAnthropicGeminiBedrock

Runtime

  • AWS EC2 on Amazon Linux
  • nginx TLS termination with auto-renewal
  • PM2, systemd services and timers
  • Docker container operations

AI Layer

  • All LLM traffic through our own gateway
  • OpenAI · Anthropic · Gemini · Bedrock
  • Call logging, usage ceilings, auto settlement
  • APM-based performance monitoring

Data

  • MariaDB with per-product instance isolation
  • Supabase Postgres, Auth and Edge Functions
  • Multi-tenant schema design

Automation

  • Self-hosted n8n workflows
  • Payments and settlement automation
  • Self-operated mail infrastructure (SPF · DKIM)
  • Daily batches and alert dispatch

If you know what needs building, we will build exactly that.

AI workflow automation, web and app development, commerce automation. You do not need a specification — we can start from deciding what to build.