Context engineering in ever changing environment - made simple
summary
An intern’s cron job that reprocesses every file hourly wastes compute when only 1% of content changes and creates ghost records when filenames or offsets shift; schema edits (adding embedding columns) force expensive reprocesses. CocoIndex is a Python-native, state-driven data pipeline that incrementally syncs sources to targets, handles CDC and deletes, and exports to a Neo4j knowledge graph via generated Cypher. The demo builds Person/Meeting/Task nodes and relationships continuously and keeps the graph up to date without downtime.
AI-written summary · about
chapters (5)
- 0:00The challenges of data pipelines
- 5:49A new programming model
- 7:53Introducing Coco Index
- 10:20Practical implementation example
- 12:35Conclusion and Q&A
related talks
12mReal-Time Data Transformation for Knowledge Graph with LLMs
25mKeynote: Think Faster: Graphs and AI for Real-Time Crisis Resolution
29mClosing the Loop: How GraphQL Gives Coding Agents Eyes on What Actually Matters
40mContext That Dreams
31mTemporal Substrate Architecture: Building Persistent Identity for Autonomous Agents
32mAgentic AI Governance: Data Foundations & Real-Time Policy Enforcement
24mIt's Time To Rethink Everything
43mAI accelerated legacy modernisation