Arize Phoenix¶

The built-in, wizard-driven source. GigaFlow reads Phoenix's spans table over a direct Postgres connection.

Prerequisites¶

A Phoenix Postgres instance reachable from the backend, and its connection string: postgresql://user:pass@host:5432/dbname.
The table name (default spans).

Connect (wizard)¶

export GIGAFLOW_BACKEND_URL=https://api.gigaflow.io/api/v1
export GIGAFLOW_API_KEY=<your key>
gigaflow setup

The wizard: creates a project → uploads the bundled arize_phoenix.yml transform → registers the datasource (source_type=arize_phoenix, source_table=spans) → runs the first sync.

Connect (API, if you prefer)¶

PID=$(curl -s -X POST "$GIGAFLOW_BACKEND_URL/projects/" -H "Authorization: Bearer $GIGAFLOW_API_KEY" \
  -H 'Content-Type: application/json' -d '{"name":"my-phoenix-project"}' | python3 -c 'import sys,json;print(json.load(sys.stdin)["project_id"])')

curl -X PUT "$GIGAFLOW_BACKEND_URL/projects/$PID/transform" -H "Authorization: Bearer $GIGAFLOW_API_KEY" \
  -H 'Content-Type: text/plain' --data-binary @gigaflow/transforms/arize_phoenix.yml

curl -X POST "$GIGAFLOW_BACKEND_URL/datasources/" -H "Authorization: Bearer $GIGAFLOW_API_KEY" \
  -H 'Content-Type: application/json' -d "{
    \"project_id\": \"$PID\",
    \"name\": \"my-phoenix\",
    \"source_type\": \"arize_phoenix\",
    \"connection_url\": \"postgresql://user:pass@host:5432/phoenix\",
    \"source_table\": \"spans\"
  }"

Transform¶

Bundled: gigaflow/transforms/arize_phoenix.yml (maps OpenInference span_kind + attributes.*). If your Phoenix uses non-standard attribute paths, copy it, edit, and pass the path during setup (or PUT .../transform).

Run¶

gigaflow sync
gigaflow compute "SELECT trace_id FROM trace_metrics WHERE run_id IS NULL"
gigaflow inspect <trace_id>