You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
# ai-agent-systems
Custom AI agent systems to streamline operations, automate tasks, and enhance decision-making. Built by a strategy-led, builder-driven team.
Agent Constitution – official guidelines covering ethics, lifecycle policies and governance for all agents.
## 🧭 Project Overview
This repository supports rapid design, development, and deployment of AI agent systems that improve operational efficiency and scale intelligently across industries.
**Use Cases Include:**
* Booking & scheduling automation
* Admin, CX, or sales workflows
* AI-powered insights dashboards
* SOP generation and team enablement
* Custom GPT-based copilots (internal or client-facing)
## 🧠 Core Team
* Chris Ponce – AI Strategist & Ops Architect
* Zach Jefferys – Full-stack AI Developer
* Mark Ehrhardt – Product Strategy & Innovation
## 🧱 System Modules
| Module | Description |
| ------------ | -------------------------------------------------------- |
| `functions/` | Server logic, APIs, agent orchestration |
| `agents/` | Modular AI agents (e.g., insights-agent, forecast-agent) |
| `frontend/` | Optional web dashboard or client UI |
| `logs/` | JSON logs for transparency and explainability |
| `sops/` | Generated SOPs or training documentation |
🔁 Agents are designed to be reusable across industries with minimal changes to input data or config files.
## 🚀 Workflow Phases
| Week | Focus | Deliverables |
| ---- | ---------------------------- | ----------------------------------- |
| 1 | Discovery & Use Case Mapping | Opportunity matrix, process audit |
| 2 | UX & Agent Design | Workflow logic, specs, data schemas |
| 3–5 | Build, Test, Integrate | Deployed agent(s), dashboards, QA |
| 6 | Handoff & Enablement | Training, SOPs, post-launch support |
## 🛠️ Tech Stack
* Node.js + Express (server-side)
* Firebase Auth + Hosting
* Stripe Checkout
* GPT-based agents (OpenAI API)
* Optional Python scripts for analytics
## 🔒 License
Apache 2.0 License
## 👣 Getting Started
```bash
# Clone the repo
git clone https://github.com/Csp-Ai/ai-agent-systems.git
# Navigate into project
cd ai-agent-systems
# Install dependencies
npm install
# Initialize Firebase hosting
npm run setup:firebase
# Deploy dashboard and hosting
npm run deploy
# Run development server
npm start
```
## 🌐 Localization & Submission Pipeline
The backend includes optional translation utilities powered by **LibreTranslate**. Pass a `locale` when calling `/run-agent` to automatically translate an agent's output. The API also exposes `/translate`, `/detect-language`, and `/locales` helpers for ad-hoc requests.
Developers can submit new agents through `/submit-agent` with metadata and a zip file. Submissions are staged for manual review before integration.
## 🔎 Governance Dashboard
The `/dashboard` directory contains a React application that connects to Firestore for real-time monitoring. To build the dashboard:
```bash
cd dashboard
npm install
npm run build
```
The compiled assets are output to `public/dashboard` and automatically served via Firebase Hosting.
## 🔥 Firebase Deployment
The `.firebaserc` already points to the production project `ai-agent-systems`. `firebase.json` now defines both `functions` and `hosting` blocks so the dashboard can be served alongside the API.
Run the deployment script to build the dashboard, copy the assets, and deploy to Firebase Hosting:
```bash
npm run deploy
```
This will output the Firebase Hosting URL on success. Deploy Cloud Functions separately if needed:
```bash
firebase deploy --only functions
```
Once deployed the public endpoints are available under:
```
https://us-central1-ai-agent-systems.cloudfunctions.net/translatehttps://us-central1-ai-agent-systems.cloudfunctions.net/reporthttps://us-central1-ai-agent-systems.cloudfunctions.net/executeAgent
```
For local development:
```bash
firebase emulators:start --only functions
# Example agent run
curl http://localhost:5001/ai-agent-systems/us-central1/executeAgent
```
If your environment blocks `firebase-public` domains and the emulator fails to start, run the server directly:
```bash
node functions/index.js
```
// rebuild trigger
About
Modular multi-agent AI architecture designed for operational automation, decision support, and governance-aware deployment. Includes structured orchestration, observability, and CI/CD integration for production-scale systems.