Skip to main content
Data Mapping

Automatic Discovery and Schema Mapping

Miradoris automatically discovers data sources, analyzes their schemas, and maps fields into the unified ontology. Manual configuration is minimized — the platform infers types, relationships, and transformations from the data itself.

95%+
Auto-mapped fields
Minutes
Time to first sync
0
Manual schema files
Schema Detection

Automatic discovery of structure and semantics

When a new data source connects, Miradoris inspects the incoming data stream to infer schema structure, field types, and semantic meaning. Column names, value distributions, and temporal patterns are analyzed to produce a candidate schema without manual intervention.

Automatic type inference from sample data
Field types are detected from value distributions and format analysis.
Semantic field classification
Column names and values are analyzed for semantic meaning.
Nested structure flattening and normalization
Complex nested objects are decomposed into flat, mappable fields.
Schema drift detection on subsequent runs
Changes in source schemas are flagged and reconciled automatically.
Schema Inspector
Field Name
Detected Type
Confidence
Mapped To
timestamp
DateTime
98%
event.timestamp
device_id
String
95%
asset.identifier
temp_c
Float
92%
sensor.temperature
lat
Float
88%
location.latitude
status_code
Integer
72%
(review)
Field Mapping
Source Fields
Ontology Properties
96% 94% 97% 91% 68%
lat
lng
device_id
temp_c
status
location.latitude
location.longitude
asset.identifier
sensor.temperature
device.status
Intelligent Mapping

Context-aware field mapping into the ontology

Detected fields are automatically matched to ontology properties using semantic similarity, naming conventions, and value-range analysis. Ambiguous mappings are surfaced for review with ranked suggestions. Approved mappings are remembered and applied to future sources.

Embedding-based semantic similarity matching
Fields are matched by meaning, not just name.
Mapping memory across similar data sources
Previous mappings inform suggestions for new sources.
Ranked suggestions for ambiguous fields
Multiple candidate mappings with confidence scores.
One-click approval and override interface
Accept or correct mappings with minimal effort.
Continuous Sync

Always current, never stale

Once mapped, data flows continuously from source to ontology. Schema changes in the source are detected and reconciled automatically. New fields are proposed for mapping, removed fields are flagged, and type changes trigger validation checks — all without interrupting the data pipeline.

Real-time incremental ingestion
Only changed records are processed and synced.
Automatic schema evolution handling
Source schema changes are detected and resolved in place.
Zero-downtime remapping on source changes
Pipeline continues uninterrupted during schema updates.
Configurable sync frequency and backpressure
Tune ingestion rate to match downstream capacity.
Sync Timeline
14:32:01 Initial sync complete
14:32:04 12,400 records/min
14:35:18 Schema change: 2 new fields
14:35:22 Auto-mapped humidity_pct
14:36:00 Sync healthy
Quality Dashboard
Completeness
98.2%
Freshness
<5s
Accuracy
99.1%
Duplicates
0.03%
Data Quality

Continuous monitoring and validation

Miradoris monitors data quality at every stage of the pipeline. Missing values, out-of-range readings, format violations, and duplicate records are detected in real time. Quality metrics are tracked per source and per field, enabling data teams to identify and resolve issues before they propagate.

Per-field quality scoring
Every field is scored for completeness, accuracy, and consistency.
Anomaly detection on ingestion streams
Outliers and unexpected patterns are flagged in real time.
Duplicate entity detection and merge
Matching records are identified and consolidated automatically.
Configurable validation rules and alert thresholds
Define acceptable ranges and notification policies per field.

Be among the first

We are looking for partners willing to test Miradoris in real operational environments. Early adopters get priority access to the platform at significantly reduced rates.

We'll review your request and follow up. No unsolicited contact.