Data cleaning that runs itself.
And tells you when something's wrong.
Connect Google Sheets, CSV, or Postgres. AstralRefine writes a cleaning plan, runs it on a schedule, and emails you the moment quality drops, so you stop spending Monday morning on the same export.
Already losing hours every week? Lock in $49/mo as a founding member →
47 on the waitlist · 50 founding seats · free forever to join
3
Connectors at launch
Sheets, CSV, Postgres
50
Founding-member seats
Hand-picked early users
$0
To join the waitlist
Full free tier at launch
If you clean the same data every week,
you're not alone.
“You get a new data export every Monday. You clean it manually. Every. Single. Week.”
Business Operator
“Your cleaning script works great, until the data format changes and nobody notices for two weeks.”
Data Analyst
“Your CRM was clean in January. It's not anymore.”
RevOps Manager
AstralRefine turns all of that into a job that runs in the background, checks its own work, and emails you the moment something looks off.
Where AstralRefine fits.
| Spreadsheets & manual cleanup | One-off Python scripts | Enterprise ETL platforms | AstralRefine | |
|---|---|---|---|---|
| Set-up time | 0 (you’re already doing it) | Days, if you can code | Weeks, with IT | ~5 minutes |
| Runs on a schedule | No | Only if you build the cron | Yes | Yes |
| Catches silent breakage | No | No | Sometimes (extra cost) | Yes, built in |
| Price | Your time | Your engineer’s time | $1,000+/mo | Free, then $49/mo |
| Made for | Anyone | Engineers | Data teams with budget | The person living with the data |
How it works
Three steps.
Then it runs forever.
01
Connect your data
Google Sheets, CSV, or Postgres. OAuth or a connection string. No setup call, no IT ticket. Takes about 90 seconds.
02
AI proposes a cleaning plan
AstralRefine reads your columns and proposes what to fix, column by column, step by step. You review it, adjust anything you want, and confirm. Nothing runs against real data until you say so.
03
It runs on a schedule
Daily, weekly, monthly, whatever matches your data cycle. If something fails or quality drops, you get an email. Otherwise, it just happens.
Features
Everything you do with data.
All in one place.
Connect once. Clean forever.
Google Sheets, CSV, and Postgres at launch. Salesforce and HubSpot next.
How connectors workKnow before anyone else does.
Every run logs quality metrics. If null rates spike or a job fails, you get an email before the dashboard does.
How alerts workA cleaning plan, written for you.
The AI reads your columns and proposes what to fix. You approve every step before anything runs.
How workflows workDo the math.
If you spend 3 hours a week cleaning the same data, at a $60/hour loaded cost, that's $9,360 a year.
AstralRefine's Pro plan is $1,188 a year.
Even if you only save half that time, the math makes the case.
A note from the founder
I built AstralRefine because I watched a colleague spend three hours every Monday cleaning the same Salesforce export. That's not a data problem, it's a time problem that's been normalised. The tools to fix it exist for engineers, but not for the people who actually live with the data. I'm fixing that.
A fair price.
Locked in for life.
Free forever to start. Paid plans begin at $49/month, a fraction of what you're already spending in wasted hours. Founding members lock in launch pricing for life.
Free
$0/mo
3 workflows · 50 runs/mo · 10k rows/run
Starter
$49/mo
20 workflows · 1,000 runs/month · all connectors
Pro
$99/mo
Unlimited workflows and runs · priority support
The questions everyone asks
Both. The free plan covers 3 workflows, 50 runs/month, and up to 10,000 rows per run, enough for most small teams. Paid plans start at $49/month when you need more.
It reads your columns and proposes a cleaning plan, what type each column is, what to standardise, what to flag. You review every step before anything runs. The AI doesn’t make changes; it makes suggestions.
It often will, on the first pass. That’s why every plan is reviewed by you before it runs, and you can preview the result on the first 100 rows of real data before committing. Nothing changes without your sign-off.
At launch: Google Sheets, CSV upload, and PostgreSQL. Salesforce, HubSpot, and AWS S3 are next. If your stack isn’t listed, tell us, the connector roadmap is built from what early users ask for.
Spreadsheets and scripts run when you remember to run them, and they fail silently when the data shape changes. AstralRefine runs on a schedule, monitors every run for null rates and row counts, and emails you the moment quality drops.
Source data is never stored, we read it to run your workflow and write back the cleaned version. Connection credentials are AES-256 encrypted in a vault and only decrypted at runtime. Disconnect a connector and the credentials are deleted.
The first 50 paying users. Founding members lock in $49/mo (or $490/year) for life, get early access before public launch, monthly advisory calls, and a permanent badge on their account. To qualify you answer 3 questions about how you’d use AstralRefine, most people on the waitlist won’t see the offer.
Public launch is June 2026. Founding members get access before then.
50 founding members. That's it.
Lock in launch pricing.
For good.
Fifty paid early supporters will lock in today's price for life , even as we ship Pro features. To qualify you answer three questions. Most people on the waitlist won't see the offer.
- Early access before public launch (June 2026)
- $49/mo (or $490/year) locked for life
- Monthly founder calls + private community