Context & Catalogue Awareness — How Fovea Understands Your Data
Fovea's intelligence starts with your data catalogue. Before answering any question, it reads your schema, business glossary, metadata classifications, and governance rules — so every AI interaction is accurate, contextual, and trusted.
- Instant dashboards - Turn raw data into actionable dashboards with natural language prompts
- Seamless automation - Automate your workflows with the help of Fovea
- Insights on-demand - Ask Fovea questions about your data and get insights instantly
Sign up for a 30-day free trial
How It Works
A Catalogue That Powers Every AI Interaction
Fovea's intelligence starts with the catalogue. Before it answers a single question, it reads your schema, business glossary, metadata classifications, and governance rules — building a complete, accurate picture of your data estate that it carries into every conversation.
Schema & Structural Awareness
Fovea automatically reads the schema of every connected data source — tables, columns, data types, primary keys, and foreign key relationships. It builds an internal map of your entire data estate so AI-generated queries are structurally correct from the first attempt.
- Auto-ingests schema on connection
- Maps cross-source relationships
- Tracks column data types and constraints
- Updates as schema evolves
Business Glossary & Semantic Layer
Business users think in terms of 'revenue', 'churn', or 'active customers' — not column names. Fovea maps these business terms to the underlying technical fields, so natural language questions resolve to the right data every time, regardless of how your columns are named.
- Map business terms to technical fields
- Supports aliases and synonyms
- Prevents ambiguous column resolution
- Consistent across all AI interactions
Column-Level Metadata & Tagging
Each column in your catalogue can carry rich metadata — descriptions, sensitivity classifications, ownership, and quality scores. Fovea reads this metadata before generating any query, ensuring it never exposes PII fields or queries deprecated columns.
- Sensitivity & PII classification
- Column descriptions and context
- Data owner attribution
- Quality score integration
Data Lineage Awareness
Fovea understands where data comes from. When you ask a question about a derived metric or transformed column, Fovea traces it back to its source, letting you trust the answer and audit the logic behind it.
- Source-to-report lineage tracking
- Transformation chain visibility
- Impact analysis on upstream changes
- Supports compliance and audits
Multi-Turn Context Memory
Every conversation with Fovea builds on the last. When you say 'break it down by region' after asking about revenue, Fovea knows exactly which dataset, filters, and time window you mean — without you having to repeat yourself.
- Preserves full conversation state
- Resolves pronouns and references
- Handles filter inheritance across turns
- Resets cleanly on new sessions
Semantic Search Across Data Assets
Don't know which table holds what you need? Search in plain language. Fovea uses semantic matching to find the most relevant tables, columns, and datasets — even when the naming doesn't match your terminology.
- Natural language asset discovery
- Fuzzy matching on column names
- Cross-source search
- Surfaces related datasets automatically
Live Catalogue Synchronisation
Your data doesn't stand still — and neither does Fovea's catalogue. Schema changes, new data sources, and updated metadata are picked up automatically, keeping the AI's knowledge of your data estate current at all times.
- Automatic schema change detection
- New source onboarding support
- Metadata refresh on schedule or trigger
- Zero manual catalogue maintenance
Cross-Source Catalogue Federation
Infoveave connects to databases, data warehouses, flat files, APIs, and streaming sources. The catalogue federates all of these into a single unified view, so Fovea can answer questions that span multiple systems without you having to know where the data lives.
- Unified view across all connectors
- Resolves cross-source joins automatically
- Single namespace for all data assets
- No manual data movement required
Governance Integration
The Catalogue Enforces What You Permit
Fovea doesn't operate outside your governance framework — it operates inside it. Every AI interaction is bounded by the rules you set in the catalogue.
Governance-Aware Queries
Fovea respects row-level security, column masking, and access policies defined in your governance layer. Catalogue metadata tells it which data a user is permitted to see — and it never generates a query that bypasses those rules.
Certified & Trusted Datasets
Data assets in the catalogue can be marked as certified. Fovea prioritises certified datasets over uncertified ones when resolving ambiguous queries, so business users always get answers from trusted data.
Deprecation & Lifecycle Management
When a column or table is deprecated in the catalogue, Fovea stops using it automatically — and can guide users toward the recommended replacement. No stale queries. No outdated answers.
Connected Capability
The Catalogue Is What Makes the AI Query Builder Accurate
When you ask "Show top customers by revenue this quarter", Fovea doesn't guess which table holds revenue or which column maps to customers. It looks it up in the catalogue — finds the right schema, applies your business glossary, checks governance rules, and generates a query that runs correctly the first time.
See how Fovea's catalogue awareness transforms data questions into trusted answers.
Frequently Asked Questions
What is a data catalogue and why does Fovea need one?
A data catalogue is a structured inventory of your data assets — tables, columns, relationships, business definitions, and metadata. Fovea uses it to understand what your data means, not just what it contains. Without it, AI-generated queries would be generic guesses. With it, every answer is grounded in your specific data model and business context.
Does Fovea automatically build the catalogue, or do I need to set it up manually?
Fovea automatically reads schema information from connected data sources on connection. You can enrich it with business glossary terms, column descriptions, sensitivity tags, and ownership metadata — but the structural layer (tables, columns, relationships) requires no manual setup.
How does Fovea handle business terms like 'revenue' that map to different columns in different systems?
The business glossary layer lets you define canonical terms and map them to the correct columns per data source. When a user asks about 'revenue', Fovea resolves it to the right field in the right table — even when different systems use different column names for the same concept.
Can Fovea access sensitive or restricted data columns?
No. Fovea respects column-level sensitivity classifications and governance policies defined in the catalogue. If a column is marked as PII or restricted, Fovea will not include it in generated queries or surface it in answers — enforcing your access rules automatically.
What happens when the schema changes — does the catalogue stay up to date?
Yes. Fovea monitors connected sources for schema changes and refreshes the catalogue automatically. New tables, added columns, or modified data types are picked up without manual intervention, keeping the AI's knowledge of your data estate current.
How does catalogue awareness improve the AI Query Builder?
The catalogue is what makes the AI Query Builder reliable. When you describe a query in plain English, Fovea looks up the schema to find the exact tables and columns, checks the business glossary to resolve terminology, applies governance rules to filter out restricted fields, and then generates SQL that runs correctly the first time.
Ready to see Infoveave in action?
