Create export - Ticksupply API

Authorizations

X-Api-Key

string

header

required

Your API key. Get one from the dashboard at https://app.ticksupply.com/api-keys

Headers

Idempotency-Key

string<uuid>

Unique key for idempotent requests. If you retry a request with the same key, you'll receive the original response without the operation being performed again.

Must be a valid UUID (any version — v4 recommended for uniqueness), up to 128 characters.

Maximum string length: 128

Body

application/json

datastream_id

integer<int64>

required

Datastream ID to export

start_time

required

Start timestamp. Accepts multiple formats:

Nanoseconds since Unix epoch (integer): 1703116800000000000
Nanoseconds since Unix epoch (string): "1703116800000000000"
ISO 8601 datetime: "2024-01-15T10:00:00Z"
ISO 8601 with fractional seconds: "2024-01-15T10:00:00.123456789Z"

Example:

1703116800000000000

end_time

required

End timestamp. Same formats as start_time.

Example:

1703203200000000000

schema

default:raw

Export column schema. Controls which columns appear in the output. Defaults to "raw" (collection_timestamp_ns + raw JSON data blob).

Accepts:

"raw" — default two-column output
A built-in schema name (e.g., "normalized") — flat columns extracted from JSON
A schema ID (e.g., "sch_0194a1b2c3d4e5f6a7b8c9d0e1f2a3b4") — custom schema by ID
An inline object with columns array — ad-hoc column definitions

"raw" is a reserved schema name — it is always accepted here but is not returned by GET /v1/export-schemas.

Example:

"normalized"

format

enum<string>

default:csv

Output file format. Defaults to csv for backwards compatibility.

Available options:

csv,

parquet

format_options

object

Optional per-format options. Shape must match format; mismatched or unknown keys produce a 400 invalid_argument response.

Server defaults are applied when keys are omitted, and the resolved options are echoed back on every export response so downstream pipelines can pin the exact configuration that was used.

Option 1
Option 2

Response

Export job accepted for processing

string

required

Prefixed export job ID (e.g., exp_0194a1b2c3d4e5f6a7b8c9d0e1f2a3b4)

Pattern: ^exp_[a-f0-9]{32}$

Example:

"exp_0194a1b2c3d4e5f6a7b8c9d0e1f2a3b4"

datastream_id

integer<int64>

required

Datastream ID being exported

Example:

123

start_time

integer<int64>

required

Start of data range (nanoseconds since Unix epoch)

Example:

1704067200000000000

end_time

integer<int64>

required

End of data range (nanoseconds since Unix epoch)

Example:

1704153600000000000

format

enum<string>

required

Output container format.

csv — gzip-compressed CSV with a header row (.csv.gz). Type information is lost; consumers parse strings.
parquet — columnar Parquet (.parquet). Preserves ClickHouse types (Decimal, DateTime64(9), Nullable, Array → LIST), supports per-column compression, and is read efficiently by pandas, Polars, DuckDB, Athena, and Spark.

Available options:

csv,

parquet

format_options

object

required

Server-resolved per-format options, echoed back with every default filled in. The shape matches format: empty object for csv, a populated ParquetFormatOptions for parquet. Pin these values in your client to keep future server default changes from affecting your pipeline.

Option 1
Option 2

status

enum<string>

required

Export job status

Available options:

queued,

running,

succeeded,

failed,

canceled

created_at

string<date-time>

required

Creation timestamp

reason

string | null

Failure reason (only present when status is "failed")

started_at

string<date-time> | null

When processing started

finished_at

string<date-time> | null

When processing finished