DeployWhisper AI

Technical Architecture & Design Document

1. Overall Project Details

DeployWhisper AI is a next-generation "Infrastructure-as-Code" (IaC) orchestration tool that leverages Natural Language Processing (NLP) to bridge the gap between developer intent and cloud reality. The platform allows developers to describe their desired infrastructure in plain English (e.g., "Deploy a scalable MERN stack to AWS with an RDS database and SSL enabled") and automatically generates, validates, and executes the necessary deployment scripts. By integrating directly with GitHub and major cloud providers, DeployWhisper eliminates the steep learning curve of Terraform and AWS Console complexities.

2. Target Audience

• Indie Hackers & Startups: Founders who want to ship fast without hiring a dedicated DevOps engineer.
• Full-Stack Developers: Engineers who know how to build apps but find cloud infrastructure configuration tedious or error-prone.
• DevOps Teams: Professionals looking to provide a "Self-Service" portal for their developers while maintaining control over architectural standards.

3. User Experience & Workflow

DeployWhisper is built on a "Conversation-to-Cloud" model. The user interacts with an AI agent that clarifies requirements before any resource is provisioned.

User Journey Flowchart

4. Technical Architecture Flow

The architecture relies on a specialized AI Inference Engine that translates natural language into structured Terraform modules, managed by a Node.js orchestration layer.

System Architecture

5. Developer Role & Implementation Focus

• AI Intent Translation: Developing the prompt engineering and vector-based logic to accurately map vague text to strict Terraform schemas.
• Real-time Streaming Logs: Implementing a WebSocket-based logging system to stream deployment progress from the runner directly to the user's dashboard.
• Secure Credential Management: Engineering a highly secure "Vault" system for managing cloud provider IAM keys and GitHub tokens.
• State Management: Leveraging React Query to handle the complex, long-running states of cloud deployments.

6. Technology Stack & Tools Used

Frontend & Dashboard:

• Core: Next.js, TypeScript, Tailwind CSS
• Editor: Monaco Editor (for IaC preview)
• State: React Query

Orchestration & Backend:

• Runtime: Node.js (API), Python (NLP/Inference)
• Infra Tooling: Terraform, AWS SDK, GitHub API
• Real-time: Socket.io

AI Infrastructure:

• LLM: Azure OpenAI (GPT-4o)
• Vector DB: Pinecone (for architectural patterns)

7. Communication Structure (Intent to Cloud)

The platform ensures a safety by verifying the generated plan before applying it to the production environment.

Provisioning Sequence Flow