AI-Driven Proposal & Media Automation System

# MORPHEUS

## AI‑Powered Proposal, Deck & Media Automation System

### Combined Phase 1 + Phase 2 (Revised Developer Brief)

---

## Project Overview

MORPHEUS is a production‑grade AI automation system designed to generate **client proposals, presentations, and optional video assets end‑to‑end**.

The system orchestrates multiple AI and creative tools through **n8n**, combining content generation, visual creation, structured presentations, intelligent decision‑making, storage, and delivery — **without email sending**.

Visual generation must rely primarily on **Freepik (including Freepik AI tools)** as the core asset and image generation source.

---

## Core Objectives

1. Generate high‑quality proposal content using AI
2. Generate and source visuals primarily via Freepik
3. Populate branded proposal templates automatically
4. Generate structured slide decks from the same source data
5. Optionally generate short AI videos for proposal summaries
6. Store and version all outputs automatically
7. Introduce an intelligence layer to guide decisions and branching logic
8. Maintain secure credential handling and production best practices

---

## Tech Stack (Fixed)

### Orchestration

* n8n (cloud or self‑hosted)

### AI & Intelligence

* OpenAI API – text and structured content generation
* Manus – autonomous reasoning / decision agent
* Grok – optional external reasoning (manual, not required for build)

### Visual & Media Generation

* Freepik (primary source for images, mockups, AI‑generated visuals)
* Canva Pro – proposal templates and layout assembly
* Gamma AI – slide deck and document generation
* Kling AI – short AI video generation

### Storage

* Google Drive (primary asset storage)

### Development

* Cursor (developer environment)

---

## System Architecture (High Level)

1. Trigger (manual or webhook)
2. Manus evaluates inputs and decides:

* Proposal type
* Required outputs (proposal, deck, video)
* Tone, structure, and visual density
3. OpenAI generates structured proposal content
4. Freepik generates or supplies visual assets
5. Canva assembles the designed proposal
6. Gamma generates a slide deck (conditional branch)
7. Kling generates a short video (conditional branch)
8. All assets are uploaded and versioned in Google Drive

---

## Phase 1 Scope – Core Automation

### 1. n8n Workflow Foundation

* Modular workflow design
* Clear node naming and logical grouping
* Environment‑variable‑driven configuration

### 2. OpenAI Integration

* API key via environment variable
* Generate:

* Proposal introduction
* Scope summary
* Key value propositions
* Closing section
* Output structured JSON for downstream use

### 3. Freepik Integration (Primary Visual Source)

Responsibilities:

* Generate or retrieve:

* AI‑generated images
* Mockups
* Background visuals
* Concept imagery
* Visual selection driven by proposal context

Integration:

* API or automated request workflow (if available)
* Manual fallback supported
* Outputs saved to Google Drive

### 4. Canva Integration

* Use provided Canva template link
* Populate placeholders such as:

* {{client_name}}
* {{proposal_date}}
* {{proposal_body}}
* Inject Freepik‑generated visuals into templates
* Export PDF or provide design link

### 5. Google Drive Integration

* Upload all generated assets into a defined folder
* Folder ID read from environment variable
* Standardized naming and versioning

---

## Phase 2 Scope – Intelligence, Decks & Media

### 6. Manus (Intelligence Layer)

Responsibilities:

* Determine proposal type (sales, pitch, internal)
* Recommend structure and tone
* Validate completeness of inputs
* Decide which outputs are required
* Influence downstream branching logic

Integration:

* Webhook or API‑based
* Called early in the workflow
* Returns structured JSON decisions

---

### 7. Gamma AI (Deck Automation)

Responsibilities:

* Convert structured proposal data into slide decks
* Generate:

* Slide titles
* Slide body content
* Narrative flow
* Output:

* Deck link and/or exported PDF/PPT

Integration:

* API or automation‑friendly interface
* Triggered conditionally based on Manus decisions

---

### 8. Kling AI (Video Generation)

Responsibilities:

* Generate short AI videos (15–60 seconds)
* Use AI‑generated scripts and style parameters
* Optional human approval checkpoint

Outputs:

* Video file or hosted link
* Stored in Google Drive

---

## Automation Logic & Branching

* Workflow branches based on:

* Proposal type
* Client tier
* Required deliverables
* Status tagging:

* Draft
* Needs Review
* Approved
* Error handling:

* Retry paths
* Graceful failure with logging

---

## Security & Architecture Rules

* No hard‑coded credentials
* All keys via environment variables or n8n credentials
* Least‑privilege permissions
* Clear documentation of required variables

---

## Testing Requirements

Developer must demonstrate:

1. Successful OpenAI content generation
2. Freepik visual generation or retrieval
3. Canva proposal generation
4. Google Drive upload and versioning
5. Manus‑driven branching logic
6. Gamma deck generation (or mocked if API‑limited)
7. Kling video generation (or mocked if API‑limited)

---

## Deliverables

* n8n workflow export(s) (JSON)
* Setup instructions
* Environment variable list
* Architecture explanation (written or Loom)
* Notes on future extensibility

---

## Ideal Developer Profile

* Proven n8n production experience
* Strong understanding of AI workflows and agents
* Experience with visual or media automation
* Architectural thinking (not just node wiring)
* Clear communication and documentation

---

## Timeline & Budget

* Estimated timeline: 10–20 days
* Budget: Open, based on experience and quality

---

## Future Phases (Not Included)

* Client portals
* Analytics and proposal scoring
* CRM integration
* Multi‑language support
* Autonomous follow‑ups