AI, Oracles, & Blockchains: Solving the Unstructured Data Problem
summary
Industry initiative with six major financial institutions uses three LLMs plus Chainlink oracles and blockchains to ingest PDFs/emails, reach automated consensus, and output an ISO-aligned unified golden record; phase one runs on Project CALM, processes simple corporate actions, and yields promising raw LLM results without fine-tuning, moving toward near‑real‑time publication and Swift MT/ISO dissemination. Reported pain: $3–5M per firm and data as root cause in 56% of errors; press release draws unusually high interest.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:48Outline of Corporate Actions Data Problem
- 1:58Balance Between Standards and Technology
- 3:44The Opportunity of AI, Blockchains, & Oracles for Corporate Actions Data
- 8:10Starting With A Problem Statement
- 10:55A Perfect Moment: Three Technologies Coming Together
- 14:04Project CALM Workflow
- 17:14The Unified Golden Record
- 20:02Integration With Existing Systems
- 21:21Key Takeaways
- 23:02What’s Next for the Corporate Actions Initiative
- 26:52How Asset Issuers Can Benefit
related talks
46mBeyond Black-Box AI: Designing Transparent Automation for Regulated Industries
22mAI and Blockchain Shaping The Future
22mStructuring the Unstructured : Empowering the AI Workflows
01h 00mComplex Data Ingestion with Open Source AI
45mAI Chats: Challenges in Business Integration
51mAvoid common LLM pitfalls
33mDeep research at scale
41mFireside Chat: Harnessing AI for Financial Services Innovation