Frequently Asked Questions
Got questions? This way.

Trust3 IQ is the universal context engine that can connect to a variety of data sources, such as relational data stores and analytical data warehouses to unify metadata, business semantics, metrics, relationships, and governance rules into a continuously updated model of business context. This unified context can then be exposed to AI agents, RAG workflows, and analytical apps via APIs and MCP to improve the accuracy and reliability of AI agents.

This eliminates manual bottlenecks, makes it scalable across domains and can easily bring this context to AI agents to improve their reliability and business outcomes. Because IQ also includes the security context, governance and compliance are embedded from the start.

Enterprise context is the shared layer of knowledge that helps AI systems understand a business the way its people do. It brings together all the elements that define how an organization operates i.e the business definitions, data relationships, governance rules, semantics, and institutional knowledge into one standardized, machine-readable framework.

Access to data alone isn’t enough, you need the context to make sense of the data. Without context, AI agents don’t know what business terms mean, which data sources to trust, or how to interpret relationships between entities. Enterprise context eliminates that ambiguity, ensuring AI agents deliver consistent, accurate, compliant, and trustworthy results across every function.

In short, it’s the critical middleware that turns raw data into actionable intelligence. It is the foundation that makes AI agents truly work for business.

To create or enrich context using Trust3 IQ, connect your data sources such as Databricks, Snowflake, and MySQL, then choose the data assets across those sources. Trust3 IQ automatically builds context from these data sources by:

1. Ingesting and mapping metadata: It extracts metadata, maps it to business concepts, identifies entities, discovers relationships, and derives metrics with high accuracy.
2. Leveraging advanced AI: IQ uses a combination of LLMs, NLP techniques, proprietary fine-tuned models, and semantic & metadata enrichment to generate context.
3. Utilizing existing data catalogs: It can augment and improve generated semantic context by incorporating information from data catalogs and semantic models (e.g., dbt, Alation, Collibra, PowerBI).
4. Integrating governance and security: IQ overlays governance and security context for use during query inference.

The following steps occur:

1. An IQ Agent orchestrates the response using multiple agents and specialized tools.
2. The semantic meaning of the query is analyzed and then contextualized to identify appropriate entities and their relationships.
3. A data analytics agent and platform-specific SQL agents are used to generate optimized SQL.
4. Finally, IQ uses a policy agent to inject the user’s context, ensuring that appropriate policies are applied before returning the results to the user.

IQ uses two key approaches for validating and maintaining the context:

1. Human Assisted Feedback Loops: Subject Matter Experts (SMEs) review and validate the generated context and SQL, enriching the generated semantics and metrics.
2. Audit Logs and Query History: IQ also leverages audit logs and query history to analyze data usage patterns to further refine the semantics and entity relationships in the context.

IQ allows you to interact with your enterprise data sources using natural language, extending the fit to a wide variety of use cases that require accessing your data & enterprise context without the overhead of translating the semantics to make sense of the data. Some common use cases are:

1. AI to BI Analytics: As demand grows across every department to “talk to the data,” or AI to BI Analytics, organizations are launching numerous AI chatbots and assistants that struggle with accuracy due to knowledge fragmentation and lack of enterprise context. Trust3 IQ eliminates this fragmentation by creating a unified semantic layer of enterprise context that can be instantly connected to any AI assistant or productivity tool. Define context once in IQ and reuse this across assistants for Strategy, FP&A, Operations, and more. Each assistant then operates with accurate, governed, and business-aware intelligence, enabling grounded conversations and faster, more reliable decisions.
2. Intelligent Sales Accelerator Agent: Sales reps spend disproportionate time chasing low-priority or stalled opportunities due to lack of timely access to historical patterns or buying signals and knowledge about sales strategies in similar accounts. With AI Agents and the enterprise context, you can make this information readily available and bring it right inside the rep workflows with recommended actions & nudges.
3. Clinical Trials Enrollment: Delays in patient enrollment, caused by time-consuming traditional methods and retrospective reporting, hinder drug development and regulatory milestones for pharma and life sciences companies. Instead, AI Agents powered by Trust3 IQ, can provide context-aware recommendations by orchestrating data and semantics across varied data and knowledge sources and help clinical operations teams to take actions to notify site managers, coordinators, and study leads with proactive, data-driven insights.
4. IQ as Data Profiling or Quality Agent: Understanding database table relationships is crucial for data quality, impact analysis, and migrations, but these connections are often hidden, making manual discovery slow and error-prone. Trust3 IQ can automate this process by analyzing a database schema and query logs to identify both direct (e.g., foreign key) and indirect (e.g., inferred or “soft”) dependencies across and within tables. Using statistical and pattern-based methods, IQ can uncover undocumented relationships, visualize dependency networks, and highlight critical or orphaned tables. This enables data engineering, governance, and analytics teams to quickly assess downstream impacts and strengthen overall data integrity.

Some of the biggest benefits with IQ are:

1. Increase AI Accuracy: Today, the biggest barrier to more widespread adoption or use of AI is accuracy and trust in the outputs of AI. IQ is engineered specifically to address this challenge and ensures that the results are accurate and grounded in context.
2. Faster time to market: IQ utilizes AI-driven metadata ingestion and semantic analysis to accelerate context extraction from your data and existing knowledge sources, thereby automating and unifying your contextual understanding. Build the context once and reuse across multiple departments and use cases, reducing the time taken to bring AI use cases to production.
3. Secure & Compliant: Because governance & security context is embedded, IQ ensures that outputs are not only accurate but also compliant and auditable.

IQ performs several read operations on the data sources it connects to with the primary purpose of understanding the various entities, their relationships and deriving business meaning. IQ does not transfer or store any of your data. IQ uses the data for:

• Metadata extraction: IQ reads the structural information about your data – table names, column names, data types, relationships, and schema definitions. This creates a map of your data landscape without accessing the actual data values.
• Data sampling: To understand the nature of your data, IQ reads small, representative samples from your tables. This typically involves reading a limited number of rows (often just a few dozens or hundreds depending on the data) to get a snapshot of actual data patterns, formats, and content types. This sampling is minimal and designed to be non-intrusive to your system’s performance.
• Data profiling: Using the sampled data, IQ performs profiling to understand what your data represents and how it behaves, enabling more accurate query interpretation. The characteristics it analyzes include:
  • Value distributions and ranges
  • Data quality indicators (nulls, duplicates, outliers)
  • Common patterns and formats
  • Statistical properties (min, max, averages, cardinality)
• Semantic analysis: IQ goes beyond technical metadata to understand the meaning of your data. This semantic layer allows IQ to interpret natural language queries correctly and map business terminology to your technical data structures. It identifies:
  • Business entities (customers, products, transactions, etc.)
  • Semantic types (email addresses, phone numbers, currency values, dates)
  • Relationships between different data elements
  • Business context and domain-specific terminology

Important Notes:

• All operations are read-only – IQ never writes to, modifies, or deletes your source data
• Data sampling is minimal to respect your system resources
• The insights gained are stored in IQ’s context engine, not your raw data
• Your actual data remains in your data sources under your control

Customer data is never used for any training purposes without consent. Please, refer to our privacy policy.

Trust3 IQ utilizes third-party LLMs for different purposes:

• Embeddings: openai/text-embedding-3-large
• Inference (semantics): anthropic/claude-sonnet-4-20250514 -OR- trust3/semantics-v1
• Inference (agent): anthropic/claude-sonnet-4-20250514

Depending on your use case & environment, you can integrate IQ in two simple ways – Ask IQ or MCP (Model Context Protocol) server.

1. Ask IQ (for instant natural language access): Ask IQ provides an out-of-the-box conversational interface that connects directly to your unified enterprise context within IQ. It allows your teams to query, explore, and analyze data in natural language. With Ask IQ, teams can “talk” to your enterprise data using the same governed, business-aware semantics defined in IQ.
2. MCP Tools (for developer and AI agent integration): IQ also exposes its unified semantic layer through Model Context Protocol (MCP) tools and APIs. These allow developers and AI platforms (like LLMs or autonomous agents) to programmatically retrieve trusted business context including metrics, relationships, and definitions and embed it directly into their workflows or decision-making systems. MCP tools make it easy to extend IQ’s semantic intelligence into any application, analytics platform, or agent framework.

Trust3 provides several deployment options to suit different customer needs.

1. Full Cloud (Public SaaS): Trust3 provides a full featured SaaS option for customers looking for fast time-to-value and minimal operational overhead. Trust3’s cloud offering is a fully SOC2 Type II compliant cloud and we take care of complete data isolation for the customers.
2. Hybrid Cloud: Trust3 hosts a lightweight control plane; while the customer teams run the runtime services entirely in their preferred cloud environment. Ideal for regulated industries, strict data-residency, or air-gapped constraints.