📊 Due Diligence Automation for Digital Assets
AI-Powered Crypto Fund Due Diligence
🚀 Project Overview
This intelligent platform automates and enhances the due diligence process for digital assets and crypto investment funds. Developed by 4DS2 students at ESPRIT University, it uses GPT-based models, data pipelines, and AI for trustworthy and scalable assessments.
From entity extraction and risk profiling to automated reports — Due_Diligence_4DS2 provides an end-to-end AI-driven solution.
🔧 Core Features
- 📡 Automated data ingestion via APIs, PDFs, GitHub, Reddit, etc.
- 🧠 Custom NLP with risk profiling using spaCy, regex, and blacklist filtering.
- 🔎 Hybrid semantic + keyword search with Qdrant & TF-IDF.
- 🤖 AI question generation using Retrieval-Augmented Generation (RAG).
- 📊 Automated PowerPoint report generation using
python-pptx.
🛠️ Tech Stack
- Languages: Python, JavaScript (React)
- ML & NLP: LLaMA 3.2, spaCy, MPNet, Cross-Encoder
- Data Pipelines: Apache Airflow, LangChain
- Retrieval: Qdrant, TF-IDF, Knowledge Graph
- Reporting: python-pptx, Seaborn
📈 Evaluation Metrics
Evaluation uses RAGAS to measure:
- Context precision, recall, and relevance
- Answer factuality and completeness
- BLEU, ROUGE, BERTScore, Exact Match
🌐 Deployment & Access
- 🚀 FastAPI backend
- 🔍 Local LLaMA inference
- 📄 SQLite for document management
📚 References
🙏 Acknowledgments
This project was developed under the guidance of the Faculty of ESPRIT University by students of 4DS2.