Tech Stack Overview
AI agents,
self-hosted
infrastructure
A multi-model agent framework running on self-hosted infrastructure — from Hermes Agent to Coolify deployment, every layer built for control and flexibility.
5 core services
running 24/7
What we run
Five services, one coherent pipeline — from agent orchestration to on-chain transactions.
Agent Framework
Hermes Agent
Multi-model, multi-platform AI agent framework from Nous Research. Orchestrates tool calls, manages sessions, connects to Telegram, Discord, and CLI.
Primary Model
Qwen 3.6 Plus
Alibaba's coding model — our daily driver for agent tasks, file operations, and multi-step reasoning.
Design Engine
GPT-Taste
Award-winning frontend skill — enforces Python-driven randomization, AIDA page structure, GSAP ScrollTriggers, and gapless bento grids.
Infrastructure
Coolify PaaS
Self-hosted platform — automatic HTTPS, zero-config deploys, full infrastructure control without cloud lock-in.
Blockchain
Hedera / GRELFswap
On-chain token swaps running on the Hedera consensus layer — fast finality, low fees, production-ready DeFi infrastructure.
How it works
From a single prompt to a deployed service — three layers, zero friction.
Agent receives the task
Hermes Agent loads the relevant skills, reads the kanban board, and determines the execution path — all autonomously.
Model generates the output
Qwen 3.6 Plus processes the request with GPT-Taste guidelines — producing production-grade HTML with GSAP animations and responsive layouts.
Coolify deploys automatically
The build artifact ships to Coolify — HTTPS provisioned, domain configured, live in seconds. No manual intervention required.
Every layer in our stack is chosen for one reason: full control over our infrastructure.
No vendor lock-in. No black boxes. Every service runs on hardware we manage.
See it live
Two sites, same pipeline. Self-hosted, HTTPS-enabled, zero cloud dependencies.
5
core services
2
live deployments
0
cloud dependencies