Know if this week's file is safe to trust.

FilePulse checks your recurring CSV export — sales, inventory, transactions, whatever you get every week — against a set of rules you control, then compares it to last time. You see exactly what's wrong and what changed before it reaches your dashboard or your client.

Get early access

Most weekly exports get eyeballed, not checked.

If you work with recurring CSV or Excel exports from Shopify, a POS system, a CRM, or an internal database, this is probably familiar: you open the file, it looks fine, you push it into the dashboard, and three days later someone asks why revenue dropped 40% overnight. Usually it's not real — it's the file.

✕ A key column has empty values it never had before

✕ The same transaction ID shows up twice

✕ A column silently changed type or disappeared

✕ Null rates on a field jumped from 1% to 20%

✕ A product category vanished, or a new one appeared

✕ This week's data is missing the last few days entirely

✕ Sales or order volume shifted in a way that looks wrong

✕ Negative values where there should never be any

Three checks, run automatically on every upload.

Each layer catches a different kind of problem. Together, they cover the things a person would catch by eye — if they had time to actually look.

Rules that are unquestionably true

Things you already know should never happen: sales can't be negative, a transaction ID can't repeat, a key column can't be empty, the data can't be a week stale. You set the rules once; every upload gets checked against them.

e.g. "flag any row where sales_usd < 0"

What changed since last week

FilePulse remembers your last upload and compares the new one against it: schema changes, null-rate jumps, row counts swinging by an unusual amount, categories that appeared or disappeared.

e.g. "missing customer_id rate: 0.4% → 8.7%"

Shifts that don't trip a hard rule

Sometimes nothing is technically "wrong" — the numbers just moved more than they should. FilePulse flags real distribution shifts (not noise) and tells you where the change came from, down to the segment.

e.g. "71% of the decline came from Store 14, Furniture category"

Anyone who has to trust a file they didn't create.

Freelance analysts & small agencies

You get a recurring export from a client with messy internal systems. A broken file that makes it into a dashboard is a credibility problem, not just a data problem. FilePulse is the check you run before you refresh anything.

Ops & finance teams at small companies

You're past spreadsheets but not big enough for a data platform or a dedicated data team. You need to know a report is trustworthy without hiring someone to babysit it every week.

Being built in the open, right now.

FilePulse is pre-launch. The core engine — the rule checks and week-over-week comparison described above — is actively being built and tested against real messy data.

Joining the waitlist gets you early access as soon as there's something to try, and a say in which checks matter most for your data. If there's a specific problem your weekly file has burned you on before, tell us — it'll likely shape what ships first.