See AES in action

Five project types, five domains. Each example is a validated AES project you can explore or use as a starting point.

ML agent-integrated

ML Model Factory

End-to-end ML pipeline: data discovery, model training, and evaluation with experiment tracking.

discover train evaluate
name: ml-model-factory
aes_version: "1.3"
skills:
  - discover
  - train
  - evaluate
View on GitHub →
Web dev-assist

SaaS Dashboard

Full-stack web application with component scaffolding, testing, and deployment automation.

scaffold test deploy
name: saas-dashboard
aes_version: "1.3"
skills:
  - scaffold
  - test
  - deploy
View on GitHub →
DevOps dev-assist

Infra Autopilot

Infrastructure automation with provisioning, deployment, and rollback capabilities.

provision deploy rollback
name: infra-autopilot
aes_version: "1.3"
skills:
  - provision
  - deploy
  - rollback
View on GitHub →
Research agent-integrated

Research Pipeline

Automated research workflow: ingest papers, analyze findings, and generate structured reports.

ingest analyze report
name: research-pipeline
aes_version: "1.3"
skills:
  - ingest
  - analyze
  - report
View on GitHub →

Two modes of operation

Dev-Assist

The agent builds the project, then steps back. You drive, the agent assists. Best for Web and DevOps projects.

Web DevOps

Agent-Integrated

The agent is embedded in the running product. It operates autonomously with identity and memory. Best for ML, Research, and Assistant projects.

ML Research Assistant

Start your own project

pipx install aes-cli && aes init