Real-Time 2.0 is engineered for high-throughput event ingestion, low-latency personalization, and large-scale triggered activations. To ensure system stability and predictable performance, specific limits and Service Level Agreements (SLAs) are in place. This document outlines the expected behavior and performance characteristics when operating Real-Time workloads.
| Category | Limit / Behavior | Notes |
|---|---|---|
| Event Tables | Up to 100 event tables per Parent Segment | Includes both batch and real-time event tables. |
| ID Stitching Keys | Up to 100 IDs can be configured for stitching; IDs must have string values | Example: email, user_id, td_client_id. |
| Unique IDs per Profile | Up to 200 unique IDs per profile | When exceeded, the oldest IDs are ejected first. See Real-Time ID Stitching Overview. |
| Event Definitions | No hard limit | Combined filter length must be ≤1000 characters. |
| Real-Time Attributes | Max 200 attributes per profile (≤500 bytes each) | Includes single, list, and counter attributes. |
| Real-Time Configuration Limit | Up to 600 combined | Single, list, counter, lookup catalog, ID stitching keys, journey count, and internally created attributes combined. Imported attributes excluded. |
| Ingestion Payload | 1–500 events per batch; 1MB per event, 5MB per batch | Applies to Streaming Ingest API and SDKs. |
| Personalization API | Response size: ≤10KB per request | Larger payloads may increase latency. |
| Ingestion Throughput | Default: 2,000 events/sec per customer | Scales horizontally to 100k+ events/sec globally. |
| Decisioning Throughput | 8,000 events/sec per customer | Scales horizontally, expected upper bound 100x per customer. |
| Triggered Activations | 8,000 activations/sec per customer | Scales horizontally, expected upper bound 100x per customer. |
| Personalization API SLA | ≤100ms (p95) response time for up to 20 attributes | Enterprise Support required for SLA guarantee. |
| Triggered Activation SLA | Up to 3 minutes for activations to be delivered | Includes evaluation, ID stitching, and activation dispatch. |
| Data Sync | Batch → Real-Time: Supported. Real-Time → Batch: Not supported | RT attributes do not sync back into batch storage. |
| Scalability | All core components scale horizontally | Very large deployments (100+ Parent Segments) should validate in PoC. |
Horizontal Scale-Out: Real-Time 2.0 automatically scales across infrastructure to manage traffic spikes. Extremely large deployments should undergo validation during a Proof of Concept (PoC) phase.
Latency Expectations: Personalization responses are nearly instantaneous, typically within 100ms (p95). Triggered activations may take up to 3 minutes end-to-end. Ingestion is designed to support sustained high throughput with linear scalability.
Data Sync Characteristics: Real-time and batch attributes are complementary, with data flowing exclusively from batch into real-time. Real-time attributes are not written back to batch storage.
Error Handling: All APIs are idempotent and retry-safe, allowing for safe reattempting of requests. Most system limits return standard HTTP codes with descriptive error messages; however, some backend limits may surface as generic errors. See the Notes and Recommendations section for details.
Customers operating at extreme scales (e.g., 100+ Parent Segments or 10M+ events/sec) should plan staged rollouts to validate performance thoroughly.
If you anticipate exceeding any of the stated limits, please contact Treasure AI Support to discuss potential configuration adjustments or scaling options.
The Personalization API SLA (≤100ms p95) applies only to requests fetching up to 20 attributes. Requests for more than 20 attributes are not covered by the SLA guarantee, regardless of the definition limit.
If the combined configuration count (single, list, counter, lookup catalog, ID stitching keys, journey count, and internally created attributes) exceeds 600, Real-Time jobs will fail with a generic error:
Oops! Something went wrong. Please try again or contact your administrator for assistance.If you encounter this, verify your configuration count and contact Treasure AI Support — the limit can be raised, though doing so may affect Personalization API response times.Certain advanced features, such as A/B testing within Real-Time Journeys and the deletion of attributes/profiles, are currently under development and may be subject to future changes.