đŦTechnical Deep Dive
Full Specifications [+]
Quick Commands
git clone https://github.com/Nour-Alhendi/AI-Merchant-Intelligence âī¸ Free2AITools Nexus Index V2.0
đŦ Index Insight
FNI V2.0 for Ai Merchant Intelligence: Semantic (S:50), Authority (A:0), Popularity (P:20), Recency (R:87), Quality (Q:50).
Verification Authority
đ What's Next?
Technical Deep Dive
AI Merchant Intelligence (Hybrid Structured + RAG)
Demo
A production-style fintech analytics assistant that answers merchant risk questions using:
- Structured analytics computed from transaction data (ground truth)
- RAG knowledge base (policies & explanations)
- LLM routing to decide which system should answer each question The system answers merchant risk questions while ensuring that all financial metrics remain deterministic and auditable.
Why this project
Key idea: In real fintech risk systems, core numbers must be deterministic and auditable.
The LLM is used for explanation and context, not for inventing metrics.
Features
Structured merchant KPI computation from CSV
- total_transactions, total_revenue, average_transaction
- refunds + refund_ratio
- chargebacks + chargeback_rate
- frauds + fraud_rate
- cross_border_ratio
Policy thresholds (from knowledge base)
- refund_ratio > 30%
- chargeback_rate > 2%
- large_transactions > 10,000
- high_cross_border > 25%
RAG pipeline using LlamaIndex + HuggingFace embeddings
- FastAPI backend
- LLM-based Hybrid Router
- Streamlit UI
- Synthetic dataset generator (~1M transactions over 5 years)
Architecture (high level)
User -> Streamlit UI -> FastAPI (/ask) -> HybridRouter
- GLOBAL_STRUCTURED: dataset-wide aggregation
- MERCHANT_STRUCTURED: merchant-specific KPIs
- RAG_ONLY: definitions/policy explanations
- HYBRID: KPIs + policy explanation
Quickstart
Create environment
conda create -n rag-project-env python=3.12 -y
conda activate rag-project-env
pip install -r requirements.txtEnvironment Variables
Create a .env file in the project root and add:
GROQ_API_KEY=your_groq_api_key_hereThe application will not start without a valid API key.
Run the Backend API
uvicorn api:app --reload --port 8001API will be available at:
Swagger Docs:
Run the Streamlit UI
streamlit run app.pyExample questions
Which merchant has the highest chargeback rate?
Is merchant m_00062 above the 2% chargeback threshold?
Explain why high cross-border ratio increases fraud risk.
Compare fraud rate vs chargeback rate.
Does merchant M122 violate policy thresholds?
đ Quick Start
conda create -n rag-project-env python=3.12 -y
conda activate rag-project-env
pip install -r requirements.txtâ ī¸ Incomplete Data
Some information about this model is not available. Use with Caution - Verify details from the original source before relying on this data.
View Original Source âđ Limitations & Considerations
- âĸ Benchmark scores may vary based on evaluation methodology and hardware configuration.
- âĸ VRAM requirements are estimates; actual usage depends on quantization and batch size.
- âĸ FNI scores are relative rankings and may change as new models are added.
- â License Unknown: Verify licensing terms before commercial use.
AI Summary: Based on GitHub metadata. Not a recommendation.
đĄī¸ Model Transparency Report
Technical metadata sourced from upstream repositories.
đ Identity & Source
- id
- gh-model--nour-alhendi--ai-merchant-intelligence
- slug
- nour-alhendi--ai-merchant-intelligence
- source
- github
- author
- Nour Alhendi
- license
- tags
- data-engineering, fastapi, fintech, llamaindex, llm, machine-learning, ml-engineering, python, rag, retrieval-augmented-generation, risk-analytics, streamlit
âī¸ Technical Specs
- architecture
- null
- params billions
- null
- context length
- null
- pipeline tag
- text-generation
đ Engagement & Metrics
- downloads
- 0
- stars
- 1
- forks
- 0
Data indexed from public sources. Updated daily.