Something wrong with a delivery? Tell us.
Support here means reaching the engineers who built your pipeline, not a ticket queue. This page covers where to send what, and what you should expect back.
What happens, and who it happens to.
- Delivery missedSame dayThe engineer on that pipeline
- Coverage droppedFlagged before you see itBatch report, then a message
- Source changed shapeRepaired, then backfilledIncluded, never re-quoted
Three inboxes, clearly split.
Something's wrong with a delivery
Missing batch, coverage drop, a field that stopped populating, or numbers that look wrong. This goes straight to the people who can fix it.
support@pyronets.comGeneral questions
How something works, what a field means, whether we support a format, anything about the service that isn't urgent.
info@pyronets.comNew sources or scope changes
Adding sites to an existing pipeline, changing cadence, or starting something new. We'll re-scope and re-quote transparently.
contactus@pyronets.comOur side of it, by situation.
A scheduled batch didn't arrive, or arrived incomplete.
What we do
We investigate immediately, tell you what happened, and give you an honest estimate for the fix. You hear from us even before it's resolved.
The batch arrived but a field's null rate jumped, or record counts moved unexpectedly.
What we do
Usually we've already caught it; the validation suite flags variance before delivery. If it reached you, we treat it as a miss on our side.
A site restructured, added protection, or moved the data.
What we do
Repairing extraction is included in a managed pipeline. We fix it and tell you the impact on the affected batches.
You need to understand a field, a value, or how something was derived.
What we do
Answered directly by whoever built that part of the pipeline.
Things worth checking first.
Not to deflect, these just make the first reply more useful.
A field is null more often than expected
Check the QA report attached to that batch. Per-field null rates are listed there. If the rate moved sharply against previous batches, it's a source change and we want to know.
Record counts dropped
Volume variance thresholds should have halted the batch before it shipped. If a reduced batch reached you, send us the batch ID and we'll trace it.
A price or value looks wrong
Every record carries its source URL and collection timestamp. Send us the record and we can show you the page it came from at the time it was collected.
You need historical data you didn't collect
Backfill is sometimes possible and sometimes isn't, it depends entirely on whether the source still exposes the history. Ask and we'll tell you honestly.
You reach one of four people.
For anything time-sensitive, use support@pyronets.com rather than an individual, it reaches all of us.
Support questions
Within one business day for general questions, and considerably faster for anything affecting a live delivery. We are a small team in one time zone, which means we're not a 24/7 desk, but it also means your message reaches an engineer rather than a queue.
No. Repairing extraction when a source changes is the core of a managed pipeline. Charging separately for it would mean billing you for the exact thing you hired us to handle.
You'll usually hear from us before you notice. If a scheduled window is going to be missed, we tell you rather than letting you find out from an empty bucket.
Yes. Delivered batches are retained for an agreed window and can be re-sent on request at no cost.
The people who built your pipeline. Pyronets is a four-person team and there is no support tier between you and them.
Not a client yet?
This page is for live pipelines. If you're still deciding, the better starting point is a scoped sample from your own sources.
Start a project