SAS to Google Cloud

Architecture

SAS in. Google Cloud out.

Deterministic parsers read the SAS estate and emit native Google Cloud code — not SAS rehosted on a GCE VM.

SAS programs → SAS2PY parser → BigQuery + Dataform + Composer

SAS

Base SAS DATA step / macros

DI Studio Jobs + mappings

EG / EM Projects + flows

Viya / CAS CASL + actions

SAS2PY Parser

Deterministic parse AI optional

Row-level parity Before cutover

SQLX emit Dataform models

DAG emit Composer

Google Cloud

BigQuery Set-based SQL

Dataform Governed models

Cloud Composer Replaces SAS Grid

Cloud Functions Non-SQL remainder

IAM + CLS Least privilege

Git repos Reproducible packs

AI is an optional add-on, off by default — the conversion runs end to end without it, air-gapped if your estate requires it.

Why Google Cloud

Serverless SQL, not another SAS box

SAS Grid capacity planning never ends

BigQuery bills the query, not the box. No cluster to size for month-end.

Macros have no Git history

Dataform models are versioned, tested, and have declared dependencies.

HIPAA and VPC controls are table stakes

The modernization lands inside your project, with IAM and column-level security from the parsed lineage.

Parser output

SAS RETAIN to a BigQuery window

A DATA step running total with a BY-group reset — emitted as a window function, not a retained variable.

SAS

/* SAS running balance */
data gold;
  set txn;
  by account_id month;
  if first.month then bal = 0;
  bal + amount;
run;

→

SAS2PY
converts

BigQuery SQL

-- RETAIN → window function
SELECT account_id, month, amount,
  SUM(amount) OVER (
    PARTITION BY account_id, month
    ORDER BY txn_date
    ROWS UNBOUNDED PRECEDING
  ) AS bal
FROM txn;

RETAIN and FIRST. become a window frame. The BY-group is a PARTITION BY.

Coverage