Datasetpublic

textql/Argo-Bench

Argo-Bench One simulated year of a New York City food-delivery company, exported to an Oracle E-Business Suite 12.2 warehouse of 235 tables and 7.54 billion rows. Leaderboard & Demo · How it works · Paper · Code An enterprise ERP you can download The data that enterprise analytics runs on is rarely public. Large companies keep their orders, payouts, ledgers and customer records in ERP systems such as Oracle E-Business Suite and SAP, extended with custom tables of… See the full description on the dataset page: https://huggingface.co/datasets/textql/Argo-Bench.

sourceHugging Facecc-by-4.0updated 9d agoView on Hugging Face

5likes2kdownloads

Dataset Card

[image]

Argo-Bench

One simulated year of a New York City food-delivery company, exported to an Oracle E-Business Suite 12.2 warehouse of 235 tables and 7.54 billion rows.

[Leaderboard & Demo](https://argo-bench.com) · [How it works](https://argo-bench.com/overview) · [Paper](https://arxiv.org/abs/2610.02122) · [Code](https://github.com/TextQLLabs/Argo-Bench)

An enterprise ERP you can download

The data that enterprise analytics runs on is rarely public. Large companies keep their orders, payouts, ledgers and customer records in ERP systems such as Oracle E-Business Suite and SAP, extended with custom tables of their own, and access to them is tightly restricted. Benchmarks therefore run on public datasets, which tend to be collections of loosely related tables rather than one system whose modules have to agree, and their answer keys are only as good as the people who wrote them. On real data, fraud that was never caught has no label, and a decision that was never taken has no outcome.

Argo-Bench simulates the business instead. The warehouse can be released whole because no one in it is real, and every answer key comes from the simulator's own state rather than from an annotator.

What's in the warehouse

—A calibrated company. 81 million orders from 17,902 real New York restaurants, delivered by 81,116 couriers over calendar 2024. The world is built from 34 public datasets and reports, and calibrated to the delivery-app economics New York City publishes each quarter and to the platforms' public filings. It absorbs a real shock: the city raised the minimum pay for delivery workers to $19.56 an hour on April 1, 2024.
—One system. 159 standard EBS tables across GL, AR, AP, OM, XLA, HZ, OKC/OKS, ZX and more, where every standard table and column exists in Oracle's EBS 12.2 data dictionary, and 76 custom XX_ tables for dispatch, courier pay, incentives, promotions, sessions and support. An order resolves into dispatch decisions, courier pay, merchant payouts and balanced general-ledger journals, and the modules reconcile. The schema was designed with three ERP consultants.
—Fraud and abuse where public evidence says it happens. Couriers who steal orders or spoof their GPS, promotion-farming rings, account takeovers, shell storefronts and diverted payouts. Each pattern is calibrated to how often honest customers share a device, an address or a card, which is what makes them hard to tell apart.
—As-of views. Eleven month schemas cut every table to what existed at the end of each month, so a task can be posed as of any month-end and graded on the months after it.

Running the benchmark

The 210 tasks cover trust and safety, marketplace operations, FP&A, accounting and growth. Agents query the warehouse, analyze it in a sandboxed Python environment, and file bans, forecasts, budgets, reported figures and dashboard data sources. The grader scores each filing against the simulation's hidden state. In the paper, the strongest of 12 models, Claude Opus 5.5, averages 59.5 points and scores 95 or more on 34.8% of tasks.

The tasks, the reference agent and its sandboxes are at github.com/TextQLLabs/Argo-Bench, which also explains how to export a run for official scoring.

Argo-Bench is built by TextQL.

Layout

data/<TABLE>/*.parquet        the rows (monthly parts 2024-MM.parquet, or part-0.parquet)
data/<TABLE>/_delta_log/      a Delta Lake log over the same files (relative paths)
tables.json                   every table: rows, files, columns with types, partition column
setup/                        loaders and month views per engine (below)

Every column has one of six types, the same in every engine:

`tables.json` type	Parquet	BigQuery	Snowflake	Spark / Delta	Trino	Iceberg
`INT64`	int64	INT64	NUMBER(19,0)	BIGINT	bigint	long
`FLOAT64`	double	FLOAT64	FLOAT	DOUBLE	double	double
`STRING`	string	STRING	VARCHAR	STRING	varchar	string
`DATETIME`	timestamp[us], no zone	DATETIME	TIMESTAMP_NTZ(6)	TIMESTAMP_NTZ	timestamp(6)	timestamp
`DATE`	date32	DATE	DATE	DATE	date	date
`BOOLEAN`	bool	BOOL	BOOLEAN	BOOLEAN	boolean	boolean

Timestamps are wall-clock times with no zone. Table and column names are upper case. Keep them that way: queries and the month views quote them.

Get the files

bash

pip install -U "huggingface_hub[hf_xet]"
hf download textql/Argo-Bench --repo-type dataset --local-dir argo-bench
# or a few tables:
hf download textql/Argo-Bench --repo-type dataset --local-dir argo-bench \
    --include "data/GL_PERIODS/*" --include "tables.json" --include "setup/*"

The SQL files under setup/ refer to the data/ directory of your copy as __DATA_ROOT__. Fill it in once for the engine you use, for example:

bash

sed -i.bak 's#__DATA_ROOT__#s3://my-bucket/argo-bench/data#g' setup/spark/register_delta.sql

Every engine below reads the same Parquet files. Loading into BigQuery or Snowflake copies them into that warehouse. Delta Lake, Iceberg, Trino and DuckDB read them in place.

DuckDB (no download needed)

bash

sed 's#__DATA_ROOT__#hf://datasets/textql/Argo-Bench/data#g' setup/duckdb/views.sql > views.sql
duckdb argo.duckdb -c ".read views.sql" -c ".read setup/duckdb/month_views.sql"
duckdb argo.duckdb -c 'SELECT COUNT(*) FROM food_delivery."XX_DISPUTES"'

Over hf:// the base views take about 2 minutes and the month views about 7 (DuckDB reads Parquet footers over HTTP as it binds each view). Only the rows a query touches are fetched. Point __DATA_ROOT__ at a local data/ directory to make both near-instant. DuckDB can also read the Delta tables: SELECT * FROM delta_scan('argo-bench/data/GL_PERIODS').

BigQuery

BigQuery loads from Google Cloud Storage, so copy data/ into a bucket first. Use a bucket in (or inside) your dataset's location. Load jobs are free. The loaded dataset holds about 880 GiB of logical storage.

bash

gcloud storage cp -r argo-bench/data gs://my-bucket/argo-bench/
RELEASE_URI=gs://my-bucket/argo-bench BQ_PROJECT=my-project BQ_DATASET=food_delivery \
    bash setup/bigquery/load.sh
BQ_PROJECT=my-project BQ_DATASET=food_delivery bash setup/bigquery/month_views.sh

load.sh creates each table with its final types and loads it with one Parquet load job. 94 tables are MONTH-partitioned on the column the month views cut on, so a month view scans only its months. JOBS=N loads N tables at a time (default 8), ONLY=A,B just those tables, and BQ_LOCATION defaults to US. With AUTHORIZE=1, month_views.sh also makes each month dataset an authorized dataset on the base, so a reader granted only food_delivery_6 cannot read past June (needs jq).

Snowflake

Create a stage named "food_delivery"."ARGO_BENCH" whose root holds data/, then run load.sql and month_views.sql in a database of your choice.

—Your copy in S3, GCS or Azure (external stage):

sql

  USE DATABASE my_db;
  CREATE SCHEMA IF NOT EXISTS "food_delivery";
  CREATE STAGE "food_delivery"."ARGO_BENCH"
    URL = 's3://my-bucket/argo-bench/' STORAGE_INTEGRATION = my_integration;

—A local download (internal stage): fill in __LOCAL_ROOT__ in setup/snowflake/put.sql, then snowsql -d my_db -f setup/snowflake/put.sql.

bash

snowsql -d my_db -f setup/snowflake/load.sql
snowsql -d my_db -f setup/snowflake/month_views.sql

(snow sql -f ... with the Snowflake CLI works too.) load.sql creates each table with its final types, then runs one COPY INTO ... MATCH_BY_COLUMN_NAME = CASE_SENSITIVE per table. The schema name is the quoted, lower-case "food_delivery".

Delta Lake (Databricks, Spark, delta-rs)

Each data/<TABLE>/ is already a Delta table. Its log names the files by relative path, so it works wherever the directory is copied: local disk, S3, GCS, ADLS, or a Databricks volume. Register the tables and add the month views:

bash

sed -i.bak 's#__DATA_ROOT__#s3://my-bucket/argo-bench/data#g' setup/spark/register_delta.sql
# Databricks: run both files in a SQL editor or notebook, in the catalog you want.
spark-sql --packages io.delta:delta-spark_2.13:4.0.0 \
  --conf spark.sql.extensions=io.delta.sql.DeltaSparkSessionExtension \
  --conf spark.sql.catalog.spark_catalog=org.apache.spark.sql.delta.catalog.DeltaCatalog \
  -f setup/spark/register_delta.sql      # then the same with -f setup/spark/month_views.sql

Delta tables with timestamps use the timestampNtz table feature (reader version 3), which Databricks Runtime 13.3+, Delta 3+, delta-rs and DuckDB support. From Python, without Spark: deltalake.DeltaTable("argo-bench/data/XX_DISPUTES").to_pyarrow_table().

Iceberg

Iceberg metadata names files by absolute path, so it is written for your copy where it lives. setup/iceberg/register.py creates each table from its Parquet schema and adds the existing files (pyiceberg add_files: manifests, statistics and a name mapping are written, and no data is copied):

bash

pip install "pyiceberg[pyarrow,sql-sqlite]"
# local copy, local SQLite catalog in ./iceberg/
python setup/iceberg/register.py --root argo-bench
# a copy in object storage, into your catalog (REST, Glue, Hive, SQL, ...)
python setup/iceberg/register.py --root s3://my-bucket/argo-bench \
    --catalog prod -P type=rest -P uri=https://catalog.example.com -P warehouse=wh

Then query food_delivery.<TABLE> from any engine attached to that catalog (Spark, Trino, Snowflake, DuckDB, PyIceberg). setup/spark/month_views.sql and setup/trino/month_views.sql create the month views in that catalog as well.

Trino

Trino reads the Delta tables in place through its Delta Lake connector:

1.Copy setup/trino/delta.properties to etc/catalog/delta.properties. Set local.location to your download, or switch to the S3/GCS lines. Restart Trino.
2.Register the tables, then add the month views:

bash

sed -i.bak 's#__DATA_ROOT__#local:///data#g' setup/trino/register_delta.sql
trino --catalog delta -f setup/trino/register_delta.sql
trino --catalog delta -f setup/trino/month_views.sql
trino --catalog delta --schema food_delivery --execute 'SELECT COUNT(*) FROM xx_disputes'

For an Iceberg catalog instead, register with setup/iceberg/register.py into a catalog Trino's Iceberg connector also uses (REST, Glue, Hive metastore, JDBC). Then run month_views.sql with --catalog set to it.

Month views (as-of datasets)

Tasks are posed as of the end of a month. The full-year tables are the base (food_delivery). Eleven sibling schemas hold views that cut every table to what existed at the end of each month:

schema	rows kept
`food_delivery_1`	before 2024-02-01
`food_delivery_2`	before 2024-03-01
`food_delivery_3`	before 2024-04-01
`food_delivery_4`	before 2024-05-01
`food_delivery_5`	before 2024-06-01
`food_delivery_6`	before 2024-07-01
`food_delivery_7`	before 2024-08-01
`food_delivery_8`	before 2024-09-01
`food_delivery_9`	before 2024-10-01
`food_delivery_10`	before 2024-11-01
`food_delivery_11`	before 2024-12-01
`food_delivery`	the whole year

A view keeps a row when its creation instant (for most tables CREATION_DATE, for some a table-specific column) falls before the cutoff. Reference tables pass through whole, a few tables are period-keyed, and a few are left out of the month schemas. Each engine's month_views file carries every table's exact rule. In BigQuery (with AUTHORIZE=1), Snowflake, Databricks Unity Catalog and Trino a view reads the base with its owner's privileges, so a role granted only one month schema can read that month and not the full year.

Tables

prefix	tables	rows
`XX_`	76	2,632,600,570
`XLA_`	7	1,323,813,636
`GL_`	14	906,523,987
`RA_`	9	869,947,258
`AR_`	13	766,451,904
`OE_`	6	581,051,178
`ZX_`	8	158,424,118
`HZ_`	12	97,461,203
`CE_`	10	95,989,462
`AP_`	12	48,567,646
`OKS_`	6	25,621,703
`OKC_`	9	17,185,555
`IBY_`	6	11,182,364
`MTL_`	9	3,130,951
`QP_`	5	136,026
`HR_`	4	53,710
`FND_`	29	17,416

<details><summary>All 235 tables</summary>

table	rows	columns	files	size
`AP_CHECKS_ALL`	5,239,635	22	1	75.1 MiB
`AP_HOLDS_ALL`	31,634	14	1	528.8 KiB
`AP_HOLD_CODES`	13	11	1	3.8 KiB
`AP_INVOICES_ALL`	6,051,283	36	1	106.9 MiB
`AP_INVOICE_DISTRIBUTIONS_ALL`	7,881,526	19	1	33.0 MiB
`AP_INVOICE_LINES_ALL`	17,470,894	17	1	120.5 MiB
`AP_INVOICE_PAYMENTS_ALL`	5,643,336	15	1	69.1 MiB
`AP_PAYMENT_SCHEDULES_ALL`	6,051,283	13	1	54.6 MiB
`AP_SUPPLIERS`	99,018	33	1	2.0 MiB
`AP_SUPPLIER_SITES_ALL`	99,018	19	1	1.1 MiB
`AP_TERMS_LINES`	3	9	1	3.1 KiB
`AP_TERMS_TL`	3	11	1	3.5 KiB
`AR_ADJUSTMENTS_ALL`	0	25	12	31.9 KiB
`AR_AGING_BUCKETS`	1	9	1	2.9 KiB
`AR_AGING_BUCKET_LINES_B`	5	11	1	3.8 KiB
`AR_AGING_BUCKET_LINES_TL`	5	9	1	3.0 KiB
`AR_BATCHES_ALL`	478	22	12	105.4 KiB
`AR_CASH_RECEIPTS_ALL`	85,702,095	39	12	741.6 MiB
`AR_CASH_RECEIPT_HISTORY_ALL`	257,303,576	25	12	1.8 GiB
`AR_DISTRIBUTIONS_ALL`	158,503,648	12	12	828.8 MiB
`AR_PAYMENT_SCHEDULES_ALL`	174,181,429	29	24	2.3 GiB
`AR_RECEIPT_CLASSES`	1	12	1	3.8 KiB
`AR_RECEIPT_METHODS`	2	9	1	3.2 KiB
`AR_RECEIVABLES_TRX_ALL`	2	12	1	4.0 KiB
`AR_RECEIVABLE_APPLICATIONS_ALL`	90,760,662	27	12	1.1 GiB
`CE_BANK_ACCOUNTS`	1	34	1	10.6 KiB
`CE_BANK_ACCT_USES_ALL`	1	21	1	6.2 KiB
`CE_GL_ACCOUNTS_CCID`	1	13	1	4.2 KiB
`CE_STATEMENT_HEADERS`	251	20	1	19.1 KiB
`CE_STATEMENT_HEADERS_INT`	0	13	1	1.6 KiB
`CE_STATEMENT_LINES`	5,395,730	19	12	41.3 MiB
`CE_STATEMENT_LINES_INTERFACE`	0	14	1	1.6 KiB
`CE_STATEMENT_RECONCILS_ALL`	90,593,471	18	12	356.9 MiB
`CE_SYSTEM_PARAMETERS`	1	16	1	5.1 KiB
`CE_TRANSACTION_CODES`	6	16	1	5.1 KiB
`FND_APPLICATION`	12	9	1	3.3 KiB
`FND_APPLICATION_TL`	12	9	1	3.2 KiB
`FND_CONCURRENT_PROGRAMS`	7	10	1	3.5 KiB
`FND_CONCURRENT_PROGRAMS_TL`	7	11	1	4.1 KiB
`FND_CONCURRENT_REQUESTS`	15,764	13	1	265.3 KiB
`FND_CURRENCIES`	68	11	1	4.4 KiB
`FND_CURRENCIES_TL`	68	9	1	3.8 KiB
`FND_FLEX_HIERARCHIES`	2	8	1	2.8 KiB
`FND_FLEX_HIERARCHIES_TL`	2	11	1	4.1 KiB
`FND_FLEX_VALIDATION_QUALIFIERS`	5	6	1	2.3 KiB
`FND_FLEX_VALUES`	351	18	1	10.9 KiB
`FND_FLEX_VALUES_TL`	351	10	1	10.3 KiB
`FND_FLEX_VALUE_HIERARCHIES`	33	11	1	3.7 KiB
`FND_FLEX_VALUE_NORM_HIERARCHY`	26	12	1	4.5 KiB
`FND_FLEX_VALUE_SETS`	7	11	1	3.9 KiB
`FND_ID_FLEXS`	1	14	1	4.7 KiB
`FND_ID_FLEX_SEGMENTS`	7	17	1	5.5 KiB
`FND_ID_FLEX_SEGMENTS_TL`	7	13	1	4.3 KiB
`FND_ID_FLEX_STRUCTURES`	1	14	1	4.6 KiB
`FND_ID_FLEX_STRUCTURES_TL`	1	12	1	4.1 KiB
`FND_LANGUAGES`	29	10	1	3.9 KiB
`FND_LOOKUP_TYPES`	74	10	1	5.2 KiB
`FND_LOOKUP_TYPES_TL`	74	12	1	7.1 KiB
`FND_LOOKUP_VALUES`	292	13	1	8.4 KiB
`FND_RESPONSIBILITY`	8	12	1	4.3 KiB
`FND_RESPONSIBILITY_TL`	8	10	1	3.5 KiB
`FND_TERRITORIES`	78	7	1	2.6 KiB
`FND_TERRITORIES_TL`	78	9	1	3.8 KiB
`FND_USER`	43	10	1	4.7 KiB
`GL_ACCOUNT_HIERARCHIES`	49,294	7	1	346.3 KiB
`GL_BALANCES`	204,547	22	1	3.2 MiB
`GL_CODE_COMBINATIONS`	10,666	19	1	74.6 KiB
`GL_IMPORT_REFERENCES`	904,430,409	17	36	5.9 GiB
`GL_INTERFACE`	0	17	1	1.8 KiB
`GL_JE_BATCHES`	3,047	15	36	213.3 KiB
`GL_JE_CATEGORIES_TL`	9	10	1	3.4 KiB
`GL_JE_HEADERS`	4,354	19	36	274.3 KiB
`GL_JE_LINES`	1,821,518	17	36	17.1 MiB
`GL_JE_SOURCES_TL`	9	15	1	4.9 KiB
`GL_LEDGERS`	1	45	1	13.8 KiB
`GL_PERIODS`	26	17	1	6.3 KiB
`GL_PERIOD_STATUSES`	104	20	1	9.2 KiB
`GL_SUMMARY_TEMPLATES`	3	21	1	7.0 KiB
`HR_ALL_ORGANIZATION_UNITS`	17,905	16	1	350.5 KiB
`HR_LOCATIONS_ALL`	17,902	17	1	464.6 KiB
`HR_OPERATING_UNITS`	1	6	1	2.0 KiB
`HR_ORGANIZATION_INFORMATION`	17,902	14	1	317.4 KiB
`HZ_CODE_ASSIGNMENTS`	28	18	1	6.1 KiB
`HZ_CONTACT_POINTS`	8,452,790	18	1	152.8 MiB
`HZ_CUSTOMER_PROFILES`	4,208,326	22	1	37.0 MiB
`HZ_CUST_ACCOUNTS`	4,208,326	14	1	40.4 MiB
`HZ_CUST_ACCOUNT_ROLES`	0	15	1	1.7 KiB
`HZ_CUST_ACCT_SITES_ALL`	18,487,230	17	1	273.8 MiB
`HZ_CUST_PROFILE_CLASSES`	1	21	1	6.1 KiB
`HZ_CUST_SITE_USES_ALL`	22,695,556	17	1	323.1 MiB
`HZ_LOCATIONS`	16,578,510	34	1	351.1 MiB
`HZ_ORGANIZATION_PROFILES`	17,931	17	1	398.1 KiB
`HZ_PARTIES`	4,307,373	31	1	121.2 MiB
`HZ_PARTY_SITES`	18,505,132	15	1	265.0 MiB
`IBY_DOCS_PAYABLE_ALL`	5,643,336	31	1	128.0 MiB
`IBY_EXTERNAL_PAYEES_ALL`	99,018	12	1	1.2 MiB
`IBY_EXT_BANK_ACCOUNTS`	98,404	18	1	1.3 MiB
`IBY_PAYMENTS_ALL`	5,239,635	34	1	86.7 MiB
`IBY_PAY_INSTRUCTIONS_ALL`	1,071	23	1	28.2 KiB
`IBY_PMT_INSTR_USES_ALL`	100,900	14	1	1.3 MiB
`MTL_CATEGORIES_B`	70	10	1	4.0 KiB
`MTL_CATEGORIES_TL`	70	9	1	3.7 KiB
`MTL_CATEGORY_SETS_B`	1	11	1	3.7 KiB
`MTL_CATEGORY_SETS_TL`	1	9	1	3.0 KiB
`MTL_ITEM_CATEGORIES`	1,037,635	13	1	1.2 MiB
`MTL_PARAMETERS`	17,903	12	1	157.7 KiB
`MTL_SYSTEM_ITEMS_B`	1,037,635	32	1	4.5 MiB
`MTL_SYSTEM_ITEMS_TL`	1,037,635	10	1	4.1 MiB
`MTL_UNITS_OF_MEASURE_TL`	1	17	1	4.8 KiB
`OE_ORDER_HEADERS_ALL`	80,908,384	41	12	2.9 GiB
`OE_ORDER_LINES_ALL`	446,720,613	24	12	4.4 GiB
`OE_ORDER_SOURCES`	40	12	1	5.7 KiB
`OE_PRICE_ADJUSTMENTS`	53,422,133	35	12	614.3 MiB
`OE_TRANSACTION_TYPES_ALL`	4	13	1	4.2 KiB
`OE_TRANSACTION_TYPES_TL`	4	13	1	4.3 KiB
`OKC_K_HEADERS_ALL_B`	1,611,303	53	1	31.9 MiB
`OKC_K_HEADERS_TL`	1,611,303	11	1	16.5 MiB
`OKC_K_ITEMS`	2,505,907	21	1	40.7 MiB
`OKC_K_LINES_B`	2,505,907	49	1	64.5 MiB
`OKC_K_LINES_TL`	2,505,907	11	1	26.5 MiB
`OKC_K_PARTY_ROLES_B`	3,222,606	17	1	72.6 MiB
`OKC_K_PARTY_ROLES_TL`	3,222,606	9	1	31.0 MiB
`OKC_STATUSES_B`	8	11	1	3.6 KiB
`OKC_STATUSES_TL`	8	11	1	3.4 KiB
`OKS_BILL_CONT_LINES`	5,907,180	33	12	31.5 MiB
`OKS_BILL_TXN_LINES`	5,907,180	18	12	27.1 MiB
`OKS_K_LINES_B`	2,505,907	29	1	47.7 MiB
`OKS_K_LINES_TL`	2,505,907	10	1	26.5 MiB
`OKS_LEVEL_ELEMENTS`	6,289,622	19	1	65.1 MiB
`OKS_STREAM_LEVELS_B`	2,505,907	24	1	45.1 MiB
`QP_LIST_HEADERS_B`	33,855	29	1	510.8 KiB
`QP_LIST_HEADERS_TL`	33,855	9	1	484.3 KiB
`QP_LIST_LINES`	34,103	33	1	512.3 KiB
`QP_PRICING_ATTRIBUTES`	34,211	23	1	458.0 KiB
`QP_QUALIFIERS`	2	24	1	7.2 KiB
`RA_BATCH_SOURCES_ALL`	3	10	1	3.6 KiB
`RA_CUSTOMER_TRX_ALL`	88,806,376	26	12	1.2 GiB
`RA_CUSTOMER_TRX_LINES_ALL`	535,185,608	20	12	5.0 GiB
`RA_CUST_TRX_LINE_GL_DIST_ALL`	245,955,258	20	12	1.8 GiB
`RA_CUST_TRX_TYPES_ALL`	4	20	1	6.1 KiB
`RA_INTERFACE_LINES_ALL`	0	18	1	2.1 KiB
`RA_TERMS_B`	3	12	1	3.9 KiB
`RA_TERMS_LINES`	3	9	1	3.2 KiB
`RA_TERMS_TL`	3	10	1	3.3 KiB
`XLA_AE_HEADERS`	166,682,629	22	13	651.2 MiB
`XLA_AE_LINES`	412,340,432	25	13	4.8 GiB
`XLA_DISTRIBUTION_LINKS`	412,340,432	7	13	1.9 GiB
`XLA_EVENTS`	166,682,629	18	13	474.9 MiB
`XLA_EVENT_TYPES_B`	5	12	1	3.9 KiB
`XLA_EVENT_TYPES_TL`	5	13	1	4.4 KiB
`XLA_TRANSACTION_ENTITIES`	165,767,504	12	13	395.8 MiB
`XX_AUDIT_TRAIL`	282,177	21	1	6.7 MiB
`XX_CAMPAIGN_AUTHORITY_EVENTS`	85,454	15	1	1.5 MiB
`XX_CAMPAIGN_BUDGET_CREDITS`	34,277	20	12	1.1 MiB
`XX_CARD_FINGERPRINTS`	7,377,322	12	1	206.1 MiB
`XX_CARD_VERIFICATIONS`	2,671,548	21	1	61.7 MiB
`XX_COURIER_ACCOUNT_EVENTS`	9,802	8	12	158.5 KiB
`XX_COURIER_APPLICATIONS`	40,167	8	12	695.3 KiB
`XX_COURIER_APP_SESSIONS`	6,075,323	9	12	75.8 MiB
`XX_COURIER_DEVICES`	115,306	6	12	951.5 KiB
`XX_COURIER_EARNINGS`	69,510,237	15	12	1.1 GiB
`XX_COURIER_IDENTITY_CHECKS`	154,540	11	12	2.6 MiB
`XX_COURIER_VIOLATIONS`	112,326	9	12	1.8 MiB
`XX_COURIER_WAITS`	21,842,232	22	12	713.9 MiB
`XX_CUSTOMER_ACCOUNT_EVENTS`	818,350	20	12	22.1 MiB
`XX_CUSTOMER_DEVICES`	9,778,766	13	1	148.5 MiB
`XX_CUSTOMER_SESSIONS`	283,206,289	16	12	4.3 GiB
`XX_DELIVERY_ASSIGNMENTS`	72,820,396	11	12	1.2 GiB
`XX_DELIVERY_ATTEMPTS`	1,971,172	17	12	31.5 MiB
`XX_DELIVERY_LEGS`	69,510,237	26	12	3.4 GiB
`XX_DEVICES`	9,394,404	13	1	260.6 MiB
`XX_DISPATCH_ASSIGNMENTS`	124,712,184	18	12	1.9 GiB
`XX_DISPUTES`	58,537	17	12	1.3 MiB
`XX_DRIVER_LOCATION_OBSERVATIONS`	134,480,113	8	12	1.3 GiB
`XX_DRIVER_PROFILES`	81,116	18	1	1.1 MiB
`XX_DRIVER_QUESTS`	145,600	16	1	1.8 MiB
`XX_DRIVER_SHIFTS`	6,102,273	15	12	118.2 MiB
`XX_ENFORCEMENT_ACTIONS`	11,065	17	1	70.8 KiB
`XX_EXPERIMENT_ASSIGNMENTS`	8,062	24	12	327.2 KiB
`XX_FRAUD_WARNINGS`	24,037	16	12	763.9 KiB
`XX_GL_INTERFACE_HIST`	492,089,977	9	36	1.6 GiB
`XX_HOTSPOT_SNAPSHOTS`	109,009	7	12	403.9 KiB
`XX_INCENTIVE_COMMITMENTS`	51,985,486	26	12	1.5 GiB
`XX_INGESTION_RUNS`	16,720	19	1	417.1 KiB
`XX_INTEGRATION_INCIDENTS`	23,873	15	12	578.3 KiB
`XX_INTERFACE_ERRORS`	2,237	20	1	60.4 KiB
`XX_IP_ADDRESSES`	2,665,687	17	1	18.5 MiB
`XX_LOGIN_ATTEMPTS`	33,214,593	15	12	660.2 MiB
`XX_MARKETPLACE_ESTIMATES`	2,605,900	22	12	38.1 MiB
`XX_MEMBERSHIP_EVENTS`	21,178,841	19	12	358.1 MiB
`XX_MEMBERSHIP_ORDER_BENEFITS`	44,815,297	19	12	698.4 MiB
`XX_MERCHANT_CLOSURES`	367,997	17	1	4.8 MiB
`XX_MERCHANT_HANDOFF_CONFIRMATIONS`	27,909,875	6	12	293.6 MiB
`XX_MERCHANT_HOLIDAY_HOURS`	87,710	14	1	662.9 KiB
`XX_MERCHANT_HOURS`	125,314	15	1	869.5 KiB
`XX_MERCHANT_INTEGRATIONS`	17,902	23	1	334.8 KiB
`XX_MERCHANT_OPS_ACTIONS`	3,444	26	12	193.6 KiB
`XX_MERCHANT_RATING_SNAPSHOTS`	5,072,584	14	12	18.2 MiB
`XX_MERCHANT_READY_REPORTS`	79,167,242	20	12	1.0 GiB
`XX_MERCHANT_REGULATORY_EVENTS`	12,273	15	1	199.0 KiB
`XX_ORDER_PREP_QUOTES`	80,908,384	16	12	1.0 GiB
`XX_ORDER_PROMISES`	71,535,421	3	12	418.7 MiB
`XX_ORDER_REJECTIONS`	427,040	18	12	8.0 MiB
`XX_ORDER_STATUS_HIST`	482,054,407	13	12	2.1 GiB
`XX_PAYMENT_AUTHS`	86,015,831	16	12	1.5 GiB
`XX_PAYMENT_INSTRUMENTS`	7,357,570	16	1	129.5 MiB
`XX_PAYOUT_PERIODS`	499	16	1	18.2 KiB
`XX_PAY_PERIOD_INCENTIVES`	53	32	1	16.8 KiB
`XX_PREP_ESTIMATE_LOG`	36,588,947	18	12	964.4 MiB
`XX_PROMOTION_ALLOCATIONS`	34,097,654	18	12	416.4 MiB
`XX_PROMOTION_CHECKOUTS`	99,173,355	11	12	1.2 GiB
`XX_PROMOTION_CONTACTS`	30,685,899	12	12	288.9 MiB
`XX_PROMOTION_DECISIONS`	13,530	22	12	552.9 KiB
`XX_PROMOTION_SUBMISSIONS`	42,666,911	8	12	439.7 MiB
`XX_PROMOTION_VARIANTS`	223	14	1	9.6 KiB
`XX_PROMO_ATTEMPTS`	35,040,794	20	12	704.4 MiB
`XX_PROMO_CODES`	34,106	20	1	786.1 KiB
`XX_QUEST_INVITATIONS`	15,762,145	14	1	243.6 MiB
`XX_QUEST_PROGRESS`	5,211,466	17	1	43.6 MiB
`XX_QUEST_WINDOWS`	145,600	14	1	2.4 MiB
`XX_REFERRALS`	508,522	15	1	13.1 MiB
`XX_REFUNDS`	1,875,438	20	12	59.2 MiB
`XX_REVIEWS`	9,521,121	16	12	166.3 MiB
`XX_SUPPLY_FORECASTS`	568,764	22	1	17.0 MiB
`XX_SUPPLY_SNAPSHOTS`	1,972,007	8	12	8.5 MiB
`XX_SUPPORT_CASES`	7,365,600	30	12	195.4 MiB
`XX_TIP_ADJUSTMENTS`	164,010	14	12	3.0 MiB
`ZX_LINES`	79,212,056	21	12	406.7 MiB
`ZX_LINES_DET_FACTORS`	79,212,056	22	12	384.6 MiB
`ZX_RATES_B`	1	18	1	5.5 KiB
`ZX_RATES_TL`	1	9	1	3.0 KiB
`ZX_REGIMES_B`	1	13	1	4.1 KiB
`ZX_REGIMES_TL`	1	9	1	3.0 KiB
`ZX_TAXES_B`	1	15	1	4.6 KiB
`ZX_TAXES_TL`	1	9	1	3.0 KiB

</details>

License

CC BY 4.0.