金融团队终于有了正经的 AI 用例——VLM 解析文档比 OCR 准得多，还能直接连下游智能体，做金融自动化的建议点开教程试试。
LlamaIndex 联合创始人 Jerry Liu 在纽约举办线下工作坊，展示如何用 AI 自动化金融文档处理流程。Logan Markewich 构建了一套完整教程，将 VLM（视觉语言模型）文档解析与模式定义、业务逻辑整合为端到端工作流。相比传统 OCR，该方法在数据提取上更准确，大幅减少人工审核需求，并能轻松对接下游智能体应用。适合投资银行、会计团队、金融 AI 初创公司及金融科技企业处理大规模消费者/监管/公共金融文书。
Yesterday we gave an in-person NYC workshop on automating your financial document ingestion pipeline...
Yesterday we gave an in-person NYC workshop on automating your financial document ingestion pipelines with AI. If you're an investment bank, accounting team, finance AI startup, fintech company, and you're processing a massive amount of consumer/regulatory/public financial paperwork at scale - you might find these resources useful! @LoganMarkewich built a full set of tutorials showing how you can stitch together VLM-enabled document parsing with schemas and business logic into an e2e workflow. This is both much more accurate than existing OCR solutions in extracting out data with much less human review, and can be easily stitched together with your downstream agent applications. Check it out: github.com/logan-markewic… If you're interested in VLM-powered doc processing, come check out LlamaParse: cloud.llamaindex.ai 💬 3 🔄 5 ❤️ 29 👀 3058 📊 13 ⚡