title	Axiom
colorFrom	gray
colorTo	blue
sdk	docker
app_port	7860
pinned	false

AXIOM-RAG: Advanced Agentic Retrieval Architecture

"It's not who I am underneath, but what I do that defines me."

AXIOM-RAG is an intelligent, high-performance Retrieval-Augmented Generation (RAG) system built with a premium,It acts as an adaptive knowledge agent, instantly routing queries between heavily grounded local documents and real-time global web search.

---

What It Does

AXIOM-RAG is a state-aware analytical engine designed for relentless fact-retrieval and direct, professional outputs.

• Dynamic Query Routing: Intelligently distinguishes between document-specific queries and general knowledge, routing to either local vector stores or DuckDuckGo web search.
• Hybrid Search Engine: Combines dense retrieval (Cohere embeddings + FAISS) with sparse keyword retrieval (BM25) for unparalleled accuracy.
• Resilient Upload Pipeline: Supports robust chunked uploads for large documents (PDF, TXT, MD) within strict serverless constraints.
• Persistent Context: Integrates SQLite-backed telemetry to maintain conversation history and agentic state across sessions.

---

Tech Stack

Intelligence Core

• LangGraph & LangChain: Orchestrating the agentic state machine and RAG pipelines.
• Groq API: Blistering-fast LLM inference for relevance grading and generation.
• Cohere: State-of-the-art vector embeddings.

###Search & Memory

• FAISS (CPU): Lightning-fast dense vector similarity search.
• Rank-BM25: Precise sparse keyword matching.
• DuckDuckGo Search: Automated fallback mechanism for external intelligence.
• SQLite: Lightweight, rock-solid session and telemetry tracking.

Backend Infrastructure

• FastAPI & Uvicorn: High-concurrency Python ASGI server capable of handling chunked streaming uploads.
• HTML/CSS/JS (Vanilla): A sleek, glassmorphism-inspired dark mode frontend with dynamic transitions.

---

How to Run

Deploying AXIOM-RAG requires minimal configuration, designed to run flawlessly on local environments or serverless platforms like Hugging Face Spaces.

1. Configure the Environment

Ensure you have your API keys ready. Create a .env file in the root directory:

GROQ_API_KEY="your_groq_api_key"
COHERE_API_KEY="your_cohere_api_key"

(Optionally include OPENAI_API_KEY if utilizing alternative models).

2. Initialize the Environment

Open a terminal in the project root and set up your virtual Python environment:

# Windows
python -m venv .venv
.venv\Scripts\activate

# Mac/Linux
python -m venv .venv
source .venv/bin/activate

pip install -r requirements.txt

3. Launch the System

Spin up the backend API and serve the static frontend:

uvicorn src.main:app --reload

The AXIOM terminal will be live at http://127.0.0.1:8000. Access the interface to begin querying the intelligence stream.

---

Author

Joel Pradham
AI/ML Engineer | Backend Developer

---

Designed and Engineered to achieve the stated outcome. No compromises.