It's rebuilt from exports every week
Someone downloads the same CSVs, cleans them, and pastes them into the same template.
Automated reporting for the reports someone rebuilds every week.
We pull the data from the systems it lives in, apply your rules, build the report, and flag the numbers that look wrong. Your team reviews the report instead of assembling it.
Teams we've built for
~3 weeks
Reporting automation, start to live
A SaaS client's recurring reports went live in about three weeks.
$5,000
One-week sprint
Rebuild one existing report and ship a first automated version.
Your repo
Code your team can run
Keep the pipeline and the runbook, or keep us on after launch.
Similar work, already live
“We'd been putting off automating our reporting for ages and they had it done in about three weeks.”
“Results get logged automatically when a run finishes now. Before that we were typing them into three different spreadsheets by hand.”
“Honestly the main thing is our inventory numbers are actually right now. It all used to be typed in by hand, now the scanners feed straight into the system.”
Founded by Colin Burke in 2018.
What reporting automation has to get right
Most recurring reports are copy and paste. Someone exports from three tools, cleans the columns, fixes the same mismatches, pastes the result into a template, and sends it. The work is predictable, so it's a good first thing to automate.
The hard part isn't the chart. It's numbers people trust: the same definitions every time, odd cases handled the way your team handles them, and a clear note when something doesn't reconcile.
Signs a report is ready to automate
The best candidates repeat on a schedule, take hours to build, and have someone who can say whether the numbers are right.
The numbers depend on who built it
Definitions live in one person's head, so two versions of the same report don't match.
Errors are found after it's sent
A broken export or a missing row shows up only when someone asks why a number moved.
The BI tool covers the easy half
Dashboards show the clean data. The adjustments, notes, and exceptions still happen in a spreadsheet.
How a reporting project runs
We automate the report you already send before we change what it shows.
- / 01
Rebuild one past report
We take a report you've already sent and rebuild it from the raw sources until every line matches.
- / 02
Write down the rules
Every definition, adjustment, and exclusion your team applies by hand becomes code you can read.
- / 03
Add checks before delivery
Totals that don't reconcile, missing data, and unusual changes hold the report and go to a person.
- / 04
Run it next to the manual version
Both versions run for a few cycles. The automated one takes over when they agree.
What you get
A reporting pipeline your team can read, run, and change.
- Connections to every system the report draws from, with access your admin controls
- Your metric definitions and adjustments written as code
- Checks that hold a report back when data is missing or doesn't reconcile
- Delivery where the report already goes: email, Slack, a shared drive, or a dashboard
- Written summaries of what changed, drafted by a model and approved by a person
- A runbook, monitoring, and a log of every run
Where AI helps in a report, and where it doesn't
The numbers come from plain code. A model helps with the parts that are written, not calculated.
Calculations stay deterministic
Totals, joins, and definitions run as ordinary code, so the same inputs always give the same numbers.
Models draft the commentary
A model can summarize what moved and why from the numbers the code produced. A person approves it before it goes out.
Reports teams automate first
Each one is narrow enough to check against reports you already have.
Weekly operations report
Volumes, backlogs, and turnaround times from your operational tools, with the outliers flagged.
Monthly client reports
One report per client from the same template and data, checked before it's sent.
Finance and KPI packs
Revenue, costs, and KPIs pulled from accounting and billing tools and reconciled against each other.
Lab and experiment results
Results logged when a run finishes, instead of typed into spreadsheets.
Three fixed-price ways to start
Start with something real users can try. Then decide if it's worth going bigger.
Sprint
1 week$5,000
Rebuild one existing report from the raw sources and ship a first automated version.
MVP
~1 monthFrom $20K
One reporting pipeline end to end: connections, rules, checks, delivery, and handover.
Scale-up
2+ months$75K–$250K
Reporting across several teams or systems, with shared definitions and monitoring.
Questions before we start
Do we need a data warehouse first?
No. We can read from your tools directly. If you already have a warehouse or a BI tool, we use it.
Will the automated report match our current numbers?
That's the first test. We rebuild a report you've already sent, and we don't switch over until the numbers match or every difference is explained.
What happens when data is missing or wrong?
The report stops, and a person gets a note that says what's missing. A wrong report doesn't go out.
Can you automate Power BI or Tableau reports?
Yes. We automate the data feeds and the steps around the dashboard, so it refreshes from your sources without manual exports or copy and paste.
Can it work with Excel and Google Sheets?
Yes. A spreadsheet can be a source, the output, or both. Many teams keep the spreadsheet they know and stop filling it in by hand.
How much does reporting automation cost?
A one-week sprint is $5,000. A full reporting pipeline usually fits the MVP tier, from $20,000. We agree the price before we start.
galactus / contactaccepting projects
Send us the report you rebuild every week.
Tell us which tools the data comes from and how long the report takes. We'll tell you how we'd automate it.