Orca Studio - for Opensearch

Orca Studio is an on-device workbench for OpenSearch and Elasticsearch. Import a data dump and search it offline, or connect to a live cluster and work against it directly. Built tablet-first, and it runs on a phone too.

WHAT IT DOES

• Import NDJSON, JSON, _bulk or CSV and search it on the device. No server, no upload, works on a plane.
• Connect to OpenSearch, Amazon OpenSearch Service or Elasticsearch 7+, using basic auth, an API key header, or AWS SigV4 request signing.
• Write Query DSL in a real editor, with syntax highlighting, one-tap formatting and worked examples for match, match_phrase, multi_match, term, range, wildcard and highlight.
• Ask in plain English and get Query DSL back to review before you run it.

WHY THIS SCORE?

Open any result and Orca explains it: the full BM25 breakdown, idf and tf and field length, in the same arithmetic and the same wording a cluster uses. On imported data the explanation is computed on the device. On a live cluster you get the cluster's own _explain output.

AGGREGATIONS AND DASHBOARDS

Group by any aggregatable field with a form, or drop into raw JSON. terms, date_histogram and histogram; count, average, sum, minimum, maximum, stats and unique count. Results render as bar, line or pie charts, or as a single number. Pin any aggregation to a dashboard that persists between sessions.

MAPPINGS

Read the mapping of any index, nested fields included. On imported data you can change a field's type or analyzer, and Orca tells you first whether that needs the index rebuilt and how many documents that will touch.

OPERATOR TOOLS

• Console: send raw requests. Every one is classified Read, Writes or Destroys data before it goes, and destructive requests are confirmed by name.
• Tail: follow an index live as documents arrive.
• Cluster: health, nodes and shards at a glance.
• Move data: copy a cluster index onto the device to search offline later, or push an NDJSON or _bulk file up to a cluster.

ALSO INCLUDED

Geo points on a map, k-NN vector search with "find similar" from any document, tap-to-filter on any table cell, nested field support across queries, aggregations, inner_hits and sorting, and export to NDJSON, JSON or CSV.

PRIVACY

No analytics. No telemetry. No ads. No accounts. There is no Orca Studio server for your data to be sent to.

Imported data is encrypted on the device with SQLCipher (AES-256) under a key held in the Android Keystore. Connection passwords, API keys and AWS credentials are sealed separately, so the profile file gives up hostnames and nothing that can authenticate. Requests go straight from your device to the cluster you configured.

Plain-English translation runs on the device by default and needs no network at all. If you add your own Anthropic or OpenAI key, a request carries your question and the index's field names and types, never a document value, and the app shows you the exact payload before the first one is sent.

SAFETY

Mark a connection read-only and it refuses every write. A connection tagged Production is read-only whatever that switch says, and cleartext http:// is called out on the connection itself.

Available in English, French, German, Spanish, Italian, Portuguese, Hindi and Indonesian.