---
license: other
language:
- en
tags:
- servicenow
- workflows
- business-rules
- agents
- benchmark
pretty_name: Cascade Bench
size_categories:
- n<1K
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train.jsonl
---

# Cascade Bench

A benchmark of ServiceNow workflow trajectories ("World of Workflows++"). Each row is a
single tool action applied to a domain, together with the seed/supporting data, the table
schemas, the business rules that can fire, and the resulting cascade of execution logs and
audit records.

## Summary

- **37 samples**, one per workflow domain (e.g. `accounts_payable_processing`,
  `incident_escalation`, `change_management`).
- One split: `train`.

## Columns

| Column | Type | Description |
|---|---|---|
| `domain` | string | Workflow domain / sample id |
| `topology` | string | Cascade topology (e.g. `linear`) |
| `total_brs_fired` | int | Business rules that actually fired |
| `expected_br_count` | int | Expected number of business rules |
| `audit_count` | int | Number of deduped audit records |
| `raw_audit_count` | int | Number of raw audit records |
| `tool_name` | string | The action invoked |
| `parameters` | string (JSON) | `{table_name, operation, fields}` — `fields` keys are domain-specific |
| `seed_data` | string (JSON) | Initial record state for the target table |
| `supporting_data` | string (JSON) | Map of related table name → rows |
| `schema` | string (JSON) | Map of table name → column schema |
| `business_rules` | list[struct] | `name, fires_on_table, filter_condition, trigger_sequence, trigger_type, order, script` |
| `ewm_logs` | list[struct] | Execution-log entries (`u_br_name`, `u_table_name`, `u_field_name`, `u_old_value`, `u_new_value`, …) |
| `audits` | list[struct] | Deduped field-level audit (`tablename, fieldname, oldvalue, newvalue, documentkey`) |
| `raw_audits` | list[struct] | Raw field-level audit, same shape as `audits` |

> The four JSON-string columns (`parameters`, `seed_data`, `supporting_data`, `schema`)
> hold dicts whose keys (table/field names) differ per domain, so they are stored as
> serialized JSON to keep a stable Arrow schema. Call `json.loads()` to use them.

## Usage

```python
from datasets import load_dataset
import json

ds = load_dataset("ServiceNow-AI/cascade_bench", split="train")

row = ds[0]
print(row["domain"], row["tool_name"])
schema = json.loads(row["schema"])          # variable-key dict
params = json.loads(row["parameters"])
brs = row["business_rules"]                  # native list of structs
```