Financial reports AI-assistant
- Category: AI Engineering & Data Science
- Objective: Introduce an AI-Assistant to support financial reports and analysis
- Project URL: Financial AI-Assistant - FinTech
Description
✨ An AI-powered assistant to retrieve and summarize critical information from company financial filings (10-Ks), built using Retrieval-Augmented Generation (RAG), Ollama local LLMs, FAISS vector database, and Streamlit.
💵 Manual reading cut by 80% (~$90/hour saved). Lower operational costs (increased net margins).
⏳ Faster client reporting (higher client satisfaction). Higher client throughput (more projects handled per analyst).
🎯 Reduced human error (lower compliance risk).
🏁 Example:
- 1 analyst analyzing 100 reports manually: ~600 hours/year → ~$54,000 cost/year.
- 1 analyst using Financial Reports Assistant: ~100 hours/year → ~$9,000 cost/year.
✅ Savings: ~$45,000/year per analyst.
Made with: Python, LangChain, Ollama, FAISS, HuggingFace, PyTorch (backend), and Streamlit.