Jedify

Jedify builds an autonomous context graph over an enterprise's warehouses, BI tools, CRMs and documents so AI agents understand metric definitions, entity relationships, permissions and company terminology. It is aimed at data teams and AI app builders whose agents hallucinate or burn tokens because they only see raw schemas.

Jedify is a New York-based startup whose platform, built on its patent-pending Semantic Fusion technology, autonomously constructs a customer-specific context graph that sits between enterprise data and the AI agents that query it. Customers connect structured, semi-structured and unstructured sources (data warehouses, databases, CRMs, financial systems, BI tools, documentation, Slack and meeting recordings), and the system infers how the organisation actually uses its data by mining query logs and extracting metric definitions from BI dashboards. The resulting graph captures metric definitions, entity relationships, lineage, permissions, business rules and domain terminology, is monitored continuously for schema and definition drift, and can be inspected and edited in natural language or SQL. On top of the graph Jedify offers Data Agents, embedded AI analytics for customer-facing products, and MCP and A2A servers so external agentic applications can pull governed context. In June 2026 the company raised a $24 million Series A led by Norwest, with Snowflake Ventures investing strategically, bringing total funding past $33 million after an $8.5 million seed in September 2023. As part of that deal Jedify integrates with Snowflake Cortex AI, Cortex Agents, CoWork and Semantic Views, automating the creation and maintenance of OSI-compliant Semantic Views. TechCrunch reported 10-20 early customers, including Kiteworks and The Weather Company, which Norwest says moved from 40-45% answer accuracy out of the box to over 85% after a refinement process. Jedify positions itself against hand-built semantic layers and metadata catalogs, and it stays model-agnostic.

Category

Data & Analytics

Pricing

Contact for pricing

Target Market

CDOs, CTOs, Data Engineers, BI Analysts, Enterprise Developers

Deployment

Cloud-first

Headquarters

New York, USA

Customers

10-20 early customers (TechCrunch, June 2026)

• ✓
  Semantic Fusion context graph

  Autonomously builds a customer-specific graph of metrics, entities, lineage, permissions and business rules from connected data and knowledge sources.

• ✓
  Continuous semantic adaptation

  Monitors structured and unstructured sources for schema and definition drift so the context graph stays current without manual upkeep.

• ✓
  Universal logic layer

  Defines business concepts such as 'active user' once and applies them across every downstream tool and agent, keeping answers consistent.

• ✓
  Data Agents

  Autonomous agents grounded in the company's context graph that reason over enterprise data and return governed, explainable answers.

• ✓
  MCP and A2A servers

  Lets external agentic applications retrieve governed context and workflows through standard agent protocols rather than bespoke integrations.

• ✓
  Snowflake Semantic Views automation

  Creates and maintains OSI-compliant Semantic Views and a registry that Snowflake Cortex Agents and CoWork use to resolve complex questions.

• ✓
  Lineage inspection and editing

  Lets users inspect decision lineage and refine definitions in natural language or SQL, making agent answers auditable and correctable.

• •
  Accurate self-serve analytics

  Business users ask questions in plain language and agents answer from governed metric definitions instead of guessing from raw column names.

• •
  Snowflake Cortex grounding

  Snowflake customers auto-generate and maintain Semantic Views so Cortex Agents and CoWork return consistent answers as the schema evolves.

• •
  Embedded customer-facing analytics

  Software companies ship AI analytics inside their product, backed by a context graph that encodes their own business logic and permissions.

• •
  Fraud and finance reporting agents

  A fintech proof of concept cited by Norwest reached 98-99% accuracy and then expanded across product, finance and leadership teams.

• •
  Cutting agent token spend

  Agents retrieve only task-relevant context from the graph instead of scanning everything, which Jedify says reduces LLM token use substantially.

Jedify publishes no list pricing and has no public pricing page; deals are sales-led, and the site lists AWS Marketplace as a procurement route. Expect enterprise contracts scoped to the number of connected sources and agents.

This page was written from 7 sources, 5 on domains other than jedify.com.

1. 1.jedify.com — jedify.comvendor
2. 2.jedify.com — semantic fusionvendor
3. 3.globenewswire.com — jedify raises 24 million in series a funding to build contex
4. 4.techcrunch.com — jedify raises 24m to help companies arm ai agents with conte
5. 5.norwest.com — jedify the missing layer in enterprise ai
6. 6.snowflake.com — jedify context graphs enterprise ai agents
7. 7.ynetnews.com — sjw944pbfe