AI Copilot

The AI Copilot API exposes programmatic access to the same assistants you use in the product: natural-language pipeline generation, config suggestions, documentation drafts, static checks on custom code, provider health, and notebook-focused helpers. Responses depend on your workspace’s enabled models and data policies.

Authorization: Bearer YOUR_API_KEY_HERE

If your workspace uses URL versioning, insert /v1 before /api in the path. See API reference.

Long-running generations may return 202 with a job_id you poll on a status endpoint where your deployment supports async mode; the examples below show a common synchronous 200 shape.

Check AI status

GET https://api.planasonix.com/api/ai/status

{
  "data": {
    "available": true,
    "providers": [
      {
        "id": "azure_openai",
        "status": "healthy",
        "models": ["gpt-4o", "gpt-4o-mini"],
        "last_error": null
      },
      {
        "id": "anthropic",
        "status": "degraded",
        "models": ["claude-3-5-sonnet-20241022"],
        "last_error": "elevated_latency_region=us-east-1"
      }
    ],
    "default_model": "gpt-4o",
    "knowledge_base_indexed_docs": 1284
  }
}

Generate pipeline from natural language

POST https://api.planasonix.com/api/ai/generate-pipeline
Content-Type: application/json

{
  "prompt": "Ingest daily CSV drops from S3 prefix s3://acme-landing/sales/ into Snowflake table analytics.raw_sales, dedupe on order_id, and convert amounts to USD using the fx_rates table.",
  "target_environment": "staging",
  "constraints": {
    "max_nodes": 24,
    "allowed_credential_types": ["aws", "database"]
  }
}

{
  "data": {
    "summary": "Four-node pipeline: S3 source → type inference → dedupe on order_id → Snowflake load with currency join",
    "definition": {
      "nodes": [
        { "id": "n1", "type": "s3_source", "config": { "prefix": "s3://acme-landing/sales/" } },
        { "id": "n2", "type": "schema_inference", "config": {} },
        { "id": "n3", "type": "dedupe", "config": { "keys": ["order_id"] } },
        {
          "id": "n4",
          "type": "snowflake_load",
          "config": { "table": "analytics.raw_sales", "write_mode": "merge" }
        }
      ],
      "edges": [
        { "from": "n1", "to": "n2" },
        { "from": "n2", "to": "n3" },
        { "from": "n3", "to": "n4" }
      ]
    },
    "warnings": [
      "Specify credential_id for S3 and Snowflake before first run."
    ]
  }
}

Enhance existing pipeline

POST https://api.planasonix.com/api/ai/enhance-pipeline
Content-Type: application/json

{
  "pipeline_id": "pl_01hqxyz",
  "prompt": "Add row-level checks: order_total must equal sum(line_items); flag negative quantities.",
  "focus": "data_quality"
}

{
  "data": {
    "pipeline_id": "pl_01hqxyz",
    "suggestions": [
      {
        "kind": "add_node",
        "after": "n_transform",
        "node": {
          "type": "assertion",
          "config": {
            "expression": "order_total = SUM(line_items.amount) OVER (PARTITION BY order_id)"
          }
        }
      }
    ],
    "rationale": "Assertions keep bad aggregates from reaching the warehouse without blocking the whole batch when configured in warn mode."
  }
}

Generate node configuration

POST https://api.planasonix.com/api/ai/generate-node-config
Content-Type: application/json

{
  "node_type": "sql",
  "intent": "Pivot monthly revenue columns into long format for BI tools",
  "sample_input_schema": {
    "columns": [
      { "name": "region", "type": "string" },
      { "name": "rev_2024_01", "type": "decimal" },
      { "name": "rev_2024_02", "type": "decimal" }
    ]
  }
}

{
  "data": {
    "node_type": "sql",
    "config": {
      "query": "SELECT region, month, revenue FROM analytics.wide_monthly UNPIVOT (revenue FOR month IN (rev_2024_01, rev_2024_02))"
    },
    "dialect": "snowflake"
  }
}

Generate documentation

POST https://api.planasonix.com/api/ai/generate-documentation
Content-Type: application/json

{
  "subject_type": "pipeline",
  "subject_id": "pl_01hqxyz",
  "audience": "data_engineers",
  "format": "markdown"
}

{
  "data": {
    "title": "Finance mart refresh",
    "markdown": "## Overview\n\nThis pipeline …\n\n## Inputs\n\n…\n\n## Failure modes\n\n…"
  }
}

Validate custom code

POST https://api.planasonix.com/api/ai/check-code
Content-Type: application/json

{
  "language": "sql",
  "dialect": "bigquery",
  "code": "SELECT customer_id, SUM(amount) AS total FROM `analytics.sales` GROUP BY 1 HAVING total > 0"
}

{
  "data": {
    "valid": true,
    "issues": [],
    "suggestions": [
      "Consider qualifying column names with table alias for readability."
    ]
  }
}

Test AI provider connection

POST https://api.planasonix.com/api/ai/test-provider
Content-Type: application/json

{
  "provider_id": "azure_openai",
  "model": "gpt-4o-mini"
}

{
  "data": {
    "provider_id": "azure_openai",
    "ok": true,
    "latency_ms": 412,
    "sample_completion": "pong"
  }
}

Parse documentation for knowledge base

POST https://api.planasonix.com/api/ai/parse-docs
Content-Type: application/json

{
  "source_url": "https://internalwiki.example.com/data/onboarding",
  "ingest_mode": "incremental",
  "tags": ["onboarding", "pipelines"]
}

{
  "data": {
    "job_id": "kbjob_01j9kd4e5f6g7h8i9",
    "chunks_indexed": 56,
    "status": "completed"
  }
}

Knowledge — feedback

POST https://api.planasonix.com/api/ai/knowledge/feedback
Content-Type: application/json

{
  "message_id": "msg_01chat_turn",
  "helpful": false,
  "reason": "cited deprecated connector",
  "correction": "Use jdbc_sap_hana instead of generic jdbc for HANA 2.0"
}

{
  "data": {
    "recorded": true,
    "feedback_id": "fbb_01j9ke5f6g7h8i9j0"
  }
}

Knowledge — stats

GET https://api.planasonix.com/api/ai/knowledge/stats

{
  "data": {
    "documents": 1284,
    "chunks": 98210,
    "last_full_reindex_at": "2025-03-25T04:00:00Z",
    "top_sources": [
      { "host": "internalwiki.example.com", "chunks": 21034 },
      { "host": "docs.planasonix.com", "chunks": 15402 }
    ]
  }
}

Knowledge — recent activity

GET https://api.planasonix.com/api/ai/knowledge/recent

{
  "data": [
    {
      "type": "ingest",
      "source": "confluence://DATA/standards",
      "status": "success",
      "at": "2025-03-27T08:12:00Z"
    }
  ]
}

Notebook — complete cell

POST https://api.planasonix.com/api/ai/notebook/complete
Content-Type: application/json

{
  "kernel": "python",
  "prefix": "import pandas as pd\n\ndf = pd.read_parquet(\"s3://acme-landing/sales/2025-03-26.parquet\")\n# TODO: compute revenue by region\n",
  "cursor_line": 4
}

{
  "data": {
    "completion": "summary = df.groupby(\"region\")[\"amount_usd\"].sum().reset_index(name=\"revenue_usd\")\nprint(summary.head())\n"
  }
}

Notebook — explain error

POST https://api.planasonix.com/api/ai/notebook/explain-error
Content-Type: application/json

{
  "kernel": "python",
  "traceback": "KeyError: 'amount_usd'",
  "cell_source": "df.groupby(\"region\")[\"amount_usd\"].sum()"
}

{
  "data": {
    "explanation": "The column amount_usd is missing from the dataframe—likely the parquet uses amount_cents instead.",
    "suggested_fix": "df[\"amount_usd\"] = df[\"amount_cents\"] / 100.0"
  }
}

Notebook — optimize

POST https://api.planasonix.com/api/ai/notebook/optimize
Content-Type: application/json

{
  "kernel": "python",
  "cell_source": "for i in range(len(df)):\n    df.at[i, 'tax'] = df.at[i, 'subtotal'] * 0.08\n",
  "goal": "vectorize for large dataframes"
}

{
  "data": {
    "optimized": "df[\"tax\"] = df[\"subtotal\"] * 0.08\n"
  }
}

Notebook — generate cell

POST https://api.planasonix.com/api/ai/notebook/generate
Content-Type: application/json

{
  "kernel": "sql",
  "dialect": "snowflake",
  "intent": "Compare this week vs last week signups by country, top 10 countries only"
}

{
  "data": {
    "cell_source": "WITH bounds AS (\n  SELECT\n    DATE_TRUNC('week', CURRENT_DATE()) AS this_week,\n    DATE_TRUNC('week', CURRENT_DATE()) - INTERVAL '7 day' AS last_week\n)\nSELECT country, …"
  }
}

AI Copilot overview

Product capabilities and guardrails.

Knowledge base

Curating sources the assistant can cite.

Overview

Data platform

Orchestration

Integration

Governance

Iceberg Catalog

Platform

Check AI status

Generate pipeline from natural language

Enhance existing pipeline

Generate node configuration

Generate documentation

Validate custom code

Test AI provider connection

Parse documentation for knowledge base

Knowledge — feedback

Knowledge — stats

Knowledge — recent activity

Notebook — complete cell

Notebook — explain error

Notebook — optimize

Notebook — generate cell

AI Copilot overview

Knowledge base

​Check AI status

​Generate pipeline from natural language

​Enhance existing pipeline

​Generate node configuration

​Generate documentation

​Validate custom code

​Test AI provider connection

​Parse documentation for knowledge base

​Knowledge — feedback

​Knowledge — stats

​Knowledge — recent activity

​Notebook — complete cell

​Notebook — explain error

​Notebook — optimize

​Notebook — generate cell

​Related topics

AI Copilot overview

Knowledge base

Check AI status

Generate pipeline from natural language

Enhance existing pipeline

Generate node configuration

Generate documentation

Validate custom code

Test AI provider connection

Parse documentation for knowledge base

Knowledge — feedback

Knowledge — stats

Knowledge — recent activity

Notebook — complete cell

Notebook — explain error

Notebook — optimize

Notebook — generate cell

Related topics