Events Semantic Data Europe 2026 RAG Accuracy Thresholds in Sustainable Finance: When “Good Enough” is Good Enough

Current achievements and expanding opportunities

For many the transition from General Knowledge LLMs to RAG architectures has been revelationary with benefits including minimal hallucinations, greater accuracy and control over sources. The vast majority of RAG use cases do not use GraphRAG but there is huge interest that continues to grow in using Graphs alongside LLMs and embedding Semantic Guardrails into AI using Ontologies, Taxonomies and other approaches.  

Using the real-world example of ‘Sustainable Finance’ an exploration of

  • When GraphRAG becomes a necessity for corporate sustainability research and benchmarking.
  • Where thresholds exist limiting the utility of modelling more data in the graph for AI workloads.
Display Date and Time
25 June, 11:40
Track
1
Cell background colour
 
Speaker Groups
Speaker
Individual Course Speaker
Events Event Speaker James Phare
Programmatic Date Range
-
Session Title
RAG Accuracy Thresholds in Sustainable Finance: When “Good Enough” is Good Enough