QueryFlow runs your ETL pipelines locally on your Mac. Visual field mapping, Python transformations, scheduled jobs, and 9 delivery destinations — for $299.99 per year instead of thousands per month.
AWS Glue is powerful. It is also extraordinarily expensive for what most data teams actually need. The dirty secret of cloud ETL is that 80% of pipelines are simple — query a warehouse, transform with Python, deliver to a destination. You do not need a serverless Spark cluster to run a 100,000-row customer sync. You need a tool that does it well and does not bill you per DPU-hour.
AWS Glue charges per Data Processing Unit per hour. Even a minimal job uses 2 DPUs at $0.44/hour. A daily ETL job that runs for 10 minutes a day costs about $4.50 per month. Sounds cheap until you realize a serious data team has 30, 50, or 100 jobs. At that scale, Glue bills of $2,000 to $10,000 per month are typical — and that is before you add the Lambda invocations, the EventBridge scheduling, the IAM role complexity, the CloudWatch logs, and the engineering time spent maintaining the whole apparatus.
QueryFlow runs ETL pipelines locally on your Mac. The scheduler executes jobs while your Mac sleeps using macOS background task APIs. When your Mac wakes, missed jobs are caught up automatically in order. The compute cost is whatever your data warehouse already charges for the query itself — there is no separate compute bill for the orchestration. The total cost of ownership for QueryFlow is $299.99 per year. Not per month. Per year.
QueryFlow replaces AWS Glue for the workloads that make up the vast majority of data engineering work in small to mid-sized teams. Daily customer syncs from Postgres to Salesforce. Hourly inventory pulls from Snowflake to Google Sheets. Weekly executive reports built in a Python notebook and delivered to an S3 bucket. Nightly CRM updates from your warehouse. These workloads do not need Spark. They need a tool that runs reliably, costs nothing per execution, and gets out of your way.
We are not going to lie to you. There are workloads where AWS Glue remains the right tool. If you are processing billions of rows per job with Spark transformations, if you have strict 24/7 SLAs that cannot tolerate a Mac being offline, or if you need to coordinate dozens of engineers around a shared data catalog, the cloud orchestrator still wins. QueryFlow is the better tool for the long tail of ETL work that does not need any of that — which, for most teams, is most of the work.
QueryFlow includes a visual ETL builder where you draw bezier curves between source and target fields, exactly like Glue's visual editor. It includes Python transformations with the same libraries Glue jobs use — pandas, numpy, pyarrow, boto3. It includes a scheduler with cron, interval, daily, weekly, and custom triggers. It delivers to S3, SFTP, email, Google Sheets, and back into any database. The capabilities are the same. The bill is not.
QueryFlow has a 14-day free trial with no credit card. Spend an afternoon recreating one of your Glue jobs in QueryFlow. Run it for two weeks. Compare the AWS bill that did not happen against the $299.99 per year QueryFlow license. The math typically reveals itself by day three.
AWS Glue is priced per Data Processing Unit (DPU) per hour. A minimum job uses 2 DPUs at roughly $0.44 per DPU-hour, so even a small daily job adds up quickly. Realistic monthly bills for a small data team run $500 to $2,000. For mid-sized teams with multiple pipelines, monthly costs of $5,000 to $10,000 are common. QueryFlow is $299.99 per year regardless of how many jobs you run — the only compute cost is whatever Snowflake, Redshift, or your source database charges for the queries themselves.
It depends on your data volume and complexity. For workloads where the SQL warehouse does most of the compute (common with Snowflake and Redshift) and your transformation logic fits in SQL or Python, QueryFlow handles it directly on a Mac. The scheduler runs jobs while the Mac sleeps and catches up automatically on wake. For truly massive volumes (multi-billion-row Spark transformations) or 24/7 always-on pipelines that cannot tolerate a Mac being offline, AWS Glue or a cloud orchestrator remains the right tool.
QueryFlow's scheduler catches up automatically on wake. When your Mac comes back online, it checks which scheduled jobs were missed and runs them in order. For most analytical workloads (daily reports, weekly syncs, hourly customer updates), this is identical in behavior to a cloud scheduler. For true 24/7 SLA-bound pipelines, you would need a server or cloud runner.
QueryFlow does not include a separate data catalog. The schema explorer reads your databases and connections directly, so the catalog is implicit — every connected database is automatically browsable. For teams that need shared cataloging across many engineers, AWS Glue Data Catalog is still useful, but for most individual data engineers and small teams the direct-connect schema browser is faster and simpler.
Yes. Flow Books support Python cells with the bundled Python 3.12 runtime and pre-installed packages: pandas, numpy, matplotlib, plotly, requests, openpyxl, pyarrow, paramiko, and boto3. SQL query results auto-inject as a pandas DataFrame. You can do exactly the kind of transformation work you would do in a Glue job, with the entire script readable in one notebook view.
Move one Glue job to QueryFlow for 14 days. Compare the AWS bill that does not happen against the $299.99/year license. The math reveals itself by day three.