Building AI Agents that actually automate Knowledge Work
summary
A practical blueprint for automating document-heavy workflows: a document MCP server composed of connectors, parser/extractors, and vector/SQL/graph indexes; an Excel pipeline that normalizes messy sheets and exposes a specialized Excel agent; benchmarks showing an LLM+tooling pipeline rising from ~70–75% baseline to 95% accuracy, exceeding a ~90% human baseline; examples include end-to-end invoice, datasheet and financial due-diligence automations that run with batch review and API integrations.
AI-written summary · about
chapters (6)
- 0:00Automating knowledge work
- 3:27Building a document toolbox
- 8:15Excel agent capabilities
- 11:11Agent design patterns
- 14:56Real-world use cases
- 17:24Conclusion
related talks
40mBuilding Agents with Agent Bricks and MCP
22mAgentic AI for Engineering Leaders | Strategy & Execution Guide
47mRevamping the Document Automation Workflow with Generative AI
36mPutting AI Agents to Work: A New Era of Open Source Connectivity with MCP
32mBuilding an Advanced Knowledge Assistant
17mBuilding Apps with Power Platform & AI
30mComposable AI: Building Next-Gen AI Agents with MCP
16mAgentic Automation - change the world of business operations?