Logo Lanfrica

GeeksonCrack/LendraAI

Domain:

socioeconomic

Record type:

software
Creator:
Gee
Host:
Intelligent financial layer for African SMEs — powered by Interswitch APIs. # LendraAI > Intelligent financial layer for African SMEs — powered by Interswitch APIs. Built for the **Interswitch × Enyata Buildathon 2025**. --- ## What is LendraAI? LendraAI is a high-fidelity, AI-powered financial platform designed to provide African SMEs with the tools they need to scale. We leverage alternative transaction data to bridge the credit gap. - **ML Credit Scoring** — Real-time risk assessment using XGBoost & SHAP explainability. - **Predictive Cash Flow** — 6-month revenue forecasting powered by LSTM neural networks. - **Loan Pre-Approval** — Instant eligibility checks based on intelligent financial profiles. - **AI Financial Advisor** — Tailored SME growth strategies powered by Llama 3 & Groq. --- ## Team (The Architects) | Name | Alias | Role | Responsibility | |------|-------|------|----------------| | **Agoro Oluwatimilehin** | Drizzy | ML Engineer | Credit Model (XGBoost + SHAP) & Interactive Dashboard UI | | **Ohine Ivori** | Lucid | Backend Engineer | Robust API Infrastructure & Interswitch Integration | | **David Akhabue** | Code | ML & Backend Engineer | Cash Flow Forecasting Service (LSTM) & Backend Integration | --- ## Tech Stack - **Machine Learning:** Python, XGBoost, SHAP, TensorFlow (LSTM), Scikit-learn - **Backend:** FastAPI, Python, Uvicorn - **Frontend:** React (Vite), Tailwind CSS (v4), Lucide Icons, Recharts - **LLM:** Groq (Llama 3.1) for AI Advisor - **APIs:** Interswitch Payment & Identity APIs - **Deployment:** Render (Backend), Vercel (Frontend) --- ## Project Structure ```text LendraAI/ ├── backend/ # FastAPI - Central API Layer ├── frontend/ # React + Vite - Concentrated Dark Dashboard ├── ml_credit/ # XGBoost Model & SHAP Explainer ├── ml_forecasting/ # LSTM Forecasting Model & Scalers ├── docs/ # Technical Documentation & Reports └── README.md ``` --- ## Quick Start ### Backend (FastAPI) ```bash cd backend pip install -r requirements.txt uvicorn main:ap …