{ "openapi": "3.0.1", "info": { "title": "RAG Pipeline Latency Profiler - SLO & Bottleneck Audit", "description": "Profile RAG traces with p50/p90/p95/p99, error rates, SLO breaches, bottleneck ranking and instrumentation gaps across arbitrary retrieval, reranking and generation stages.", "version": "0.0", "x-build-id": "sqqdxkBq3bHkoQloL" }, "servers": [ { "url": "https://api.apify.com/v2" } ], "paths": { "/acts/quanmatrix~rag-pipeline-latency-profiler/run-sync-get-dataset-items": { "post": { "operationId": "run-sync-get-dataset-items-quanmatrix-rag-pipeline-latency-profiler", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for its completion, and returns Actor's dataset items in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } }, "/acts/quanmatrix~rag-pipeline-latency-profiler/runs": { "post": { "operationId": "runs-sync-quanmatrix-rag-pipeline-latency-profiler", "x-openai-isConsequential": false, "summary": "Executes an Actor and returns information about the initiated run in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/runsResponseSchema" } } } } } } }, "/acts/quanmatrix~rag-pipeline-latency-profiler/run-sync": { "post": { "operationId": "run-sync-quanmatrix-rag-pipeline-latency-profiler", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for completion, and returns the OUTPUT from Key-value store in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } } }, "components": { "schemas": { "inputSchema": { "type": "object", "required": [ "rag_logs" ], "properties": { "rag_logs": { "title": "RAG latency spans", "minItems": 1, "maxItems": 20000, "type": "array", "description": "Each row can be an object or JSON string with at least stage and latency_ms.", "items": { "type": "object" } }, "expected_stages": { "title": "Expected pipeline stages", "uniqueItems": true, "type": "array", "description": "Used to flag missing instrumentation. Arbitrary stage names remain supported.", "items": { "type": "string", "minLength": 1 }, "default": [ "retrieval", "embedding", "reranking", "generation" ] }, "end_to_end_sla_ms": { "title": "End-to-end SLO in milliseconds", "minimum": 0, "type": "number", "description": "Optional latency target. Measured breach rate requires request_id on spans." }, "stage_sla_ms": { "title": "Per-stage SLOs", "type": "object", "description": "Optional mapping from stage name to latency target in milliseconds." }, "previous_profile": { "title": "Previous latency profile", "type": "object", "description": "Optional prior Actor output for release-regression comparison.", "default": {} }, "monthly_requests": { "title": "Monthly RAG requests", "minimum": 0, "type": "integer", "description": "Projected monthly request volume for economic impact estimates.", "default": 0 }, "cost_per_request_second_usd": { "title": "Cost per request-second (USD)", "minimum": 0, "type": "number", "description": "Optional effective infrastructure/provider cost per request-second.", "default": 0 }, "warn_sla_breach_rate": { "title": "Warn SLA breach rate", "minimum": 0, "maximum": 1, "type": "number", "description": "Warn SLA breach rate. Configure this optional control for the current audit when needed.", "default": 0.05 }, "block_sla_breach_rate": { "title": "Block SLA breach rate", "minimum": 0, "maximum": 1, "type": "number", "description": "Block SLA breach rate. Configure this optional control for the current audit when needed.", "default": 0.25 } } }, "runsResponseSchema": { "type": "object", "properties": { "data": { "type": "object", "properties": { "id": { "type": "string" }, "actId": { "type": "string" }, "userId": { "type": "string" }, "startedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "finishedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "status": { "type": "string", "example": "READY" }, "meta": { "type": "object", "properties": { "origin": { "type": "string", "example": "API" }, "userAgent": { "type": "string" } } }, "stats": { "type": "object", "properties": { "inputBodyLen": { "type": "integer", "example": 2000 }, "rebootCount": { "type": "integer", "example": 0 }, "restartCount": { "type": "integer", "example": 0 }, "resurrectCount": { "type": "integer", "example": 0 }, "computeUnits": { "type": "integer", "example": 0 } } }, "options": { "type": "object", "properties": { "build": { "type": "string", "example": "latest" }, "timeoutSecs": { "type": "integer", "example": 300 }, "memoryMbytes": { "type": "integer", "example": 1024 }, "diskMbytes": { "type": "integer", "example": 2048 } } }, "buildId": { "type": "string" }, "defaultKeyValueStoreId": { "type": "string" }, "defaultDatasetId": { "type": "string" }, "defaultRequestQueueId": { "type": "string" }, "buildNumber": { "type": "string", "example": "1.0.0" }, "containerUrl": { "type": "string" }, "usage": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "integer", "example": 1 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } }, "usageTotalUsd": { "type": "number", "example": 0.00005 }, "usageUsd": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "number", "example": 0.00005 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } } } } } } } } }