🧠 PlanSmart: Enterprise RAG & Knowledge Engine

Executive Summary: A production-grade Retrieval-Augmented Generation (RAG) system built as the centralized "AI Brain" for PlanSmart Agency. It autonomously ingests, vectorizes, and retrieves proprietary business frameworks, internal documentation, and client histories to augment Claude's reasoning capabilities via the Model Context Protocol (MCP).

💼 The Business Case & ROI (Internal "Dogfooding")

This repository is not a proof-of-concept; it is the active knowledge infrastructure powering PlanSmart's daily operations.

Core Operational Bottlenecks Solved:

• Zero-Friction Knowledge Transfer: Eliminates information silos. New team members or AI agents can instantly query historical client data, Standard Operating Procedures (SOPs), and strategic frameworks without interrupting senior staff.

• Hallucination-Free AI: By strictly grounding Claude's responses in our localized ChromaDB vector store, we ensure all generated strategies and internal outputs are 100% aligned with PlanSmart's business standards.

• Scalable Agency Infrastructure: Reduces research and data-retrieval time per project from hours to milliseconds, dramatically lowering operational overhead.

Related MCP server: Chroma MCP Server

🏗️ Architecture & Data Flow

The system employs a strict Domain-Driven Design (DDD), separating data ingestion from the retrieval mechanism to ensure scalability and ease of testing.

graph TD;
    subgraph Data Ingestion Pipeline
        A[Internal PDFs / SOPs / Markdown] -->|Chunking| B(Document Loader)
        B -->|Embedding| C[(ChromaDB Vector Store)]
    end

    subgraph Retrieval & Generation
        D[User / Agent Query] -->|Semantic Search| C
        C -->|Top-K Context Windows| E{Context Injector}
        E -->|Augmented Prompt| F[Claude 3.5 Sonnet API]
        F --> G[Validated Business Output]
    end