Jedify tested a production benchmark using context graphs instead of injecting a full schema to scope only relevant business entities before SQL generation, achieving 87% accuracy across 200 graded runs.
Public source
Publisher name
Public post
Cost gets the headlines. Accuracy is the bigger problem. Research on LLMs has documented what's known as the "lost in the middle" effect. The longer and more cluttered a…
Company
Jedify
Jedify is the context graph for enterprise AI.
- Industry
- Software Development
- Location
- New York, US
- Company size
- 11–50 employees
About Jedify
Powered by its proprietary Semantic Fusion™ technology, Jedify autonomously builds a live, customer-specific context graph on top of an enterprise’s existing data and knowledge infrastructure. By connecting structured operational data from data warehouses, SORs, CRMs, financial systems, BI tools with unstructured knowledge from file repositories, documents, playbooks, Jira, Notion, Slack, and meeting recordings, Jedify creates an AI-ready layer that continuously learns, improves, and reflects how the business actually works.
See moreLatest activity
Latest activity from Jedify
3 signals
Products & Services
Jedify announced that its Context Graph keeps the business context current and renders it straight into governed Snowflake Semantic Views.
Research & Knowledge
Jedify published a research paper testing a different approach for AI cost efficiency by pre-encoding business logic and selecting only relevant entities in context graphs.
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
MadeAi
MadeAi tested its graph-inference capability through expert validation, reviewing 7 chart and graph formats, 81 graph-inferred data points, and achieving approximately 98% accuracy.
Research & Knowledge
UnifyApps
UnifyApps cites Snowflake research findings from a Forbes piece showing that adding a context layer to data agents improved accuracy by 20% while reducing tool calls by 39%
Research & Knowledge
Neo4j
Neo4j published research from Newcastle University showing that agents delivering more than 2x precision and recall, 80% improvement in truthfulness, and 69% better answer relevancy versus vector-only agents using a Wikipedia-style graph
Research & Knowledge
Perplexity
Perplexity's Search API took the top three spots in a benchmark of Search APIs for AI agents covering 18 products across 8 providers, with all three compute settings scoring higher than every other provider tested.
Research & Knowledge
Collibra