Nothing to fall behind
Freshness stops being a pipeline metric, because there is no pipeline.
Real-time Analytics
Events read while they are still current.
Real-time analysis fails for a mundane reason: the analysis system sees a copy, and the copy is old.
A threshold checked against data that is still being written.
Windows and aggregates over data that is still being written.
Reads run against the operational layer while writes continue, so a window, a rate or a threshold is computed from the current data. There is no second pipeline, which is also why there is nothing to fall behind. The same access path serves records, documents and time-ordered data.
A threshold checked against data that is still being written.
Environment
Live writes
The operational system keeps writing while the question is being asked.
Time series · SQL
Two estates, drawn.
The same shapes, held two ways. The difference is not the storage, it is where the agreement between them lives.
Separate systems Copies kept in step
One layer One plan · one contract
What this workload moves.
Every shape below is a surface of the layer, not a format to be converted into one. They are read and written together.
- Time series Measurements and events in time order
- SQL Records, keys and joins
- Events Operational events as they happen
How the workload reaches the data.
The models this workload names, the path a request takes through them, and the surfaces that speak to the layer.
Real-time Analytics · architecture
Workload
The shape of the work itself.
- Ingest
- Aggregate
- Alert
- Explore
Data models
The models this workload names.
- Time series
- SQL
- Events
The layer
One path from a request to the data it names.
- Planning Predicates narrow the work before it runs
- Execution Rows, fields and windows answered together
- Transactions Readers and writers do not block each other
Surfaces
How the workload reaches the layer.
- SQL surface The query language the layer is documented in
- Applications Services and jobs reading and writing as they run
- Analytics & AI clients The same layer, the same access path
The work, stage by stage.
Choose a stage to read what happens there, or a shape to see every stage that handles it. Nothing on the map is a private interface.
PLOMID · Real-time Analytics ingest · aggregate · alert · explore Select a stage
Stage
Ingest
Operational writes land in the layer
Time series · SQL
01
Ingest
Operational writes land in the layer
- Time series
- SQL
02
Aggregate
Windows, rates and thresholds computed in place
- Time series
03
Alert
Conditions evaluated against the same reads
- Events
- SQL
04
Explore
The result is queried like any other data
- SQL
What changes when the data is in one place.
Stated as properties of the system rather than as results we cannot measure for you.
Thresholds on live data
Conditions are evaluated against the data as written rather than against a copy.
One definition of a rate
The same window definition serves the dashboard and the report.
Results stay queryable
An alert is a record in the layer, so it can be read against what produced it.
What this is built on.
Real-time work is decided by read consistency under write load. These are the parts that provide it.
- One data layer → Rows, documents and time-ordered events live in one system, so a question is asked once instead of once per store.
- Snapshot reads under continuous writes → Readers and writers do not block each other, which is what makes a telemetry feed and an application share one system.
- One plan per request → A request is planned once against the layer rather than per system, which is where cross-system glue usually accumulates.
- Time-ordered storage → Measurements and events are stored beside the records and documents they describe, so history and current state agree.
Workloads that sit beside this one.
Where this shows up.
Derived from the industry pages that reference this solution, so the two directions of the relationship always agree.
- Robotics → Robot state, motion traces, task outcomes and fleet coordination.
- Utilities → Grid topology, meter reads, asset condition and customer records.
- Banking → Accounts, transactions, counterparties and the relationships between them.
- Payments → Authorisations, clearing, settlement, routing and fraud signals.
- Fintech → Ledgers, balances, product events and the relationships behind them.
- Risk & Fraud → Signals, relationships, cases and decisions across an institution’s data.
- Retail → Transactions, stock, stores, promotions and customer behaviour.
- E-commerce → Orders, catalogues, inventory, sessions and fulfilment events.
- Logistics → Shipments, routes, positions, scans and delivery events.
- Telecom → Network topology, performance counters, subscriber records and incidents.
- Scientific Research → Instrument output, simulation results, reference material and datasets.
- IoT → Device fleets, telemetry, device state and the events between them.
- Cybersecurity → Telemetry, identity, asset relationships and the cases that follow.
- Observability → Metrics, logs, traces and the service records they describe.
- Autonomous Systems → Sensor streams, perception output, decisions, interventions and logs.
- Edge Infrastructure → Compute and data close to the source, with a central record behind it.