Platform > Data Layer > Calculation + Aggregation
Calculation + aggregation.
Compute metrics per asset, per site, or across an entire fleet - sum, average, count + aggregate in real-time, as part of any workflow.
CAPABILITY OVERVIEW
Real-time metrics without a separate analytics layer,
Rayven's calculation + aggregation nodes compute metrics from raw data as it flows through the platform - no data warehouse, no separate analytics layer, no scheduled batch job.
Aggregate per-UID for entity-level metrics, per-Label for site or fleet-wide views, or across a time window for trend analysis.
Every calculated value is immediately available to dashboards, AI models, workflow logic + external systems.
Inbound triggers include:
-
Cassandra time-series data (per-UID + per-Node)
-
Primary + Secondary Table records
-
Real-time incoming workflow payloads
-
Historical data over configurable time windows
Outbound triggers include:
-
Calculated metric values for dashboard KPI widgets
-
Aggregated results for AI model inputs
-
Threshold-evaluated outputs for alert triggers
-
Aggregated datasets for API endpoints or FTP export
KEY CAPABILITIES
What Calculation + Aggregation gives you.
Calculate Value for UID
Compute sum, average, minimum, maximum, count or last value for any data field across a defined time window - for a single entity UID. Runs per-asset, giving each entity its own independent calculated metric within the same workflow execution.
Calculate Value for Label
Aggregate metrics across all UIDs sharing a Label - site total, fleet average, regional maximum. A single node calculates fleet or site-wide metrics in real-time, feeding summary dashboards or comparative views across groups.
Data Summary node
Generate summary statistics across a defined time interval - totals, averages, highs + lows per period. Used for creating hourly, daily or weekly summary records from high-frequency time-series data, written directly to a Secondary Table for trend analysis.
Count Value by Time
Count the frequency of events within a configurable time window. Used for tracking how many times a condition occurred, how many transactions processed, or how many alerts fired in a period. Feeds rate-based dashboards + frequency threshold logic.
Advanced Function node
A prebuilt formula library for complex calculations: rolling averages, delta calculations, conditional formulas + multi-variable expressions. Combines with Calculate Value nodes for multi-step metric computation within a single workflow chain.
Real-time + historical aggregation
Aggregate over a rolling window of the most recent events, a fixed historical period, or a combination of both. Supports sub-minute windows for high-frequency operational metrics alongside longer-term trend calculations in the same pipeline.
HOW IT CONNECTS: EXPLAINER
Where Calculation + Aggregation fits in the Rayven Platform stack.
Calculation + aggregation nodes sit in the Data Layer, operating on data stored in Cassandra + MySQL by the real-time processing engine.
-
Raw time-series data arrives from the Integration Layer via sensors, APIs + file uploads.
-
Calculation nodes aggregate + compute metrics within the workflow.
-
Calculated values pass directly to the Execution Layer for threshold evaluation, AI inputs + automated actions.
-
The Presentation Layer receives aggregated metrics for dashboards, KPI widgets + reports.
-
Aggregated outputs can also be written back to Secondary Tables or exposed via API endpoints.
USE CASES
How Calculation + Aggregation gets used.
Fleet performance monitoring for a mining operator
A mining operator runs Calculate Value for UID across a fleet of 500 haul trucks, computing average payload, fuel consumption + cycle count per asset over rolling 24-hour windows. Fleet averages are calculated using Calculate Value for Label. A dashboard surfaces per-truck and fleet-wide KPIs, updating every 30 seconds.
Real-time revenue aggregation for a financial services platform
Transaction events arrive via API in real-time. A Calculate Value node sums revenue per customer UID over a rolling 30-day window. A Label-level aggregation calculates total revenue by advisor team. Both metrics feed a live BI dashboard - no data warehouse, no scheduled export.
Partner building a multi-client analytics platform
An MSP configures per-Label aggregation workflows for each client. Each client's data is aggregated independently using their Label assignment. Calculated metrics feed branded client dashboards - delivered as the partner's own analytics product with no visible Rayven infrastructure.
Rayven Calculation + Aggregation FAQs:
What aggregation functions are available in Rayven?
SUM, COUNT, AVG, MIN, MAX, FIRST and LAST across configurable time windows (hourly, daily, weekly, custom). Aggregations run against Cassandra time-series data, indexed by UID and timestamp. See SQL + Cassandra Storage.
Can KPIs be calculated across multiple assets simultaneously?
Yes. Aggregate nodes can loop across all UIDs in a dataset and calculate per-entity or cross-entity KPIs in a single workflow run. Results can be written to a Secondary Table or pushed directly to a dashboard. See Unified Data Tables.
How are time window aggregations configured?
Time windows are set per aggregation node - last N minutes, last N hours, calendar day, week, month or custom epoch ranges. Multiple windows can run in parallel within the same workflow. See Real-time Data Processing.
Can calculations reference multiple data sources?
Yes. A calculation workflow can pull from Cassandra (time-series), MySQL (structured tables) and external APIs simultaneously. The JavaScript Node or Formula Builder merges the inputs for the final calculation. Explore Data Transformation.
How frequently do aggregations run?
Aggregation workflows run on any schedule - every minute, hourly, daily - or trigger from an inbound event. Real-time KPI calculations can run on every data arrival. See Workflows + Triggers.
Can aggregated results feed alerts?
Yes. A calculation workflow can include a Conditional Filter after the aggregation step. When a calculated KPI breaches a threshold, an alert is fired immediately. See Notifications + Alerts.
Can I build custom formula logic beyond standard aggregation functions?
Yes. The JavaScript Node and Advanced Function Node support arbitrary calculation logic - weighted averages, rolling windows, percentile calculations and exponential smoothing. See Data Transformation.
Can aggregated data feed AI models?
Yes. Pre-calculated aggregates are a common training input for Rayven's Python ML modeller. Rolling averages, daily totals and anomaly scores all serve as feature inputs. See AI Models + Training.
Where are aggregated results stored?
Results can be written to any Rayven Secondary Table, returned as a GET API endpoint response, written to Cassandra as a new time-series metric, or displayed directly in a dashboard widget. See Unified Data Tables.
Can aggregations run across historical data retrospectively?
Yes. Any Cassandra dataset retained within the configured window is queryable retrospectively. A workflow can be configured to recalculate historical KPIs across any retained time range. See SQL + Cassandra Storage.
Also in the Data Layer:
Unified Data Tables
Structured Primary + Secondary Tables for entity records, metadata + relational data alongside Cassandra time-series.
Data Management
Configure retention policies, inspect workflow payloads, export raw data + manage data lifecycle across the platform.
Data Transformation
JavaScript, Advanced Function + Combine Data nodes for schema mapping, enrichment + normalisation within workflow processing chains.
File Parsing
Ingest + parse files from FTP, S3 + manual uploads into structured, real-time data available to workflows and AI models.
Real-time Data Processing
Sub-second ingestion + processing of live sensor, device + event data with built-in deduplication + schema validation.
AI Models + Training
Train Python ML models on Cassandra time-series data + deploy predictions as real-time workflow steps.
SQL + Cassandra Data Storage
Hybrid storage architecture - MySQL for relational records, Cassandra for time-series + event data.
Join the Shift
Discover the easy way to do something new.
Book a free 30 minute assessment with our team and we'll scope your project, needs + what a solution might look like.