01
The sender fixes the error
The supplier submits and sees the verdict. Corrections happen where the data originates, by the only party who knows what the value should be. Your team reviews results instead of chasing files.
Data submission portal
Your counterparties submit. You receive clean, auditable data.
DataOnboard collects recurring files from hundreds of external suppliers, checks each delivery against that supplier's own format, maps it to your target schema and sends errors back to the sender before anything reaches your systems.
Free to start. No credit card.
Intake queue
October 2026 · 200 counterparties
128
Validated
15
Received
9
Blocked
48
Not yet submitted
| Counterparty | Format | Date | Status |
|---|---|---|---|
| Alder Property Services | xlsx v3 | 02 Oct | validated |
| Meridian Fund Administration | csv v1 | 02 Oct | validated |
| Castell & Vane | xlsx v2 | 03 Oct | blocked |
amount is 12× the expected range in 41 rows. Annual values in a monthly column? Confirm or resubmit. |
|||
| Northgate Asset Services | export v1 | 03 Oct | received |
| Harbour Street Management | xlsx v1 | 04 Oct | validated |
| Pellworth Advisory | csv v4 | 04 Oct | blocked |
unit_id missing in 3 rows (lines 118, 240, 241). Required by the target schema. |
|||
| Sato & Lund | csv v2 | 06 Oct | validated |
| Greyfield Holdings | – | – | not yet submitted |
01 The situation
Two hundred counterparties. Two hundred formats. One deadline.
A property manager sends an Excel export from their system. A fund administrator sends a CSV with its own column names. A supplier sends last month's file with two columns renamed and the amounts now monthly instead of annual.
Someone on your team merges all of it into one schema by hand. Errors show up after import, where they cost the most to fix. Who has delivered and who has not lives in a side spreadsheet. When the auditor asks where a number came from, the answer is a person's memory.
Onboarding one new counterparty format takes about a day of that person's time. In a corpus of seven real counterparty files we studied, none parsed with standard tooling. Every file was its own structure.
02 How it works
One deterministic path from upload to export
-
STEP 01
Define what you need.
Describe your target schema and the rules a valid record must pass. One schema per data type, shared by every counterparty.
-
STEP 02
Your counterparties upload what they already produce.
Each supplier's format is captured once as a data contract. From then on they upload the same file they always sent. Nothing to configure on their side.
-
STEP 03
Every delivery is validated, mapped and reconciled.
The contract is parsed, the mapping runs as generated SQL, your rules and cross-file reconciliations execute on the result. The same delivery produces the same output every time.
-
STEP 04
Errors go back to the sender. Clean data goes to you.
A delivery that fails gets a field-level verdict in plain language, addressed to the person who can fix it. A delivery that passes waits for your approval and exports to your target format, API or warehouse.
03 What is different
Built for data that has to survive an audit
02
Every number traces back to a source cell
Mapping and validation run as reproducible SQL on each delivery. You can replay any past delivery and show how each field was derived. There is no model in the path between the file and your records.
03
"Cannot be mapped" is an answer, too
When a file does not contain what your schema needs, the sender gets a specific explanation of what is missing. A parse error never reaches a supplier.
04
You see the whole wave
One queue per period shows who has delivered, what is validated, what is blocked and why, and who is still outstanding.
05
Set up a format once, accept repeats automatically
The expensive step is understanding a counterparty's format the first time. After that, repeat deliveries in the same structure flow through on their own.
Create your free account
Free to start. No credit card.
04 Who uses it
Any organisation that receives recurring data from outside
The portal does not know what a tenancy schedule, a GP report or a supplier questionnaire is. It knows your schema, each counterparty's format and the rules between them. The vertical is configuration.
Real estate funds and asset managers
Tenancy schedules and accounting files from dozens or hundreds of property managers, every month, into fund accounting. In production today
Fund and investor reporting
Periodic data from many GPs, administrators or portfolio companies into one reporting model.
ESG and supplier data
Metrics and questionnaire answers from thousands of suppliers, collected against one schema with validation at submission.
Insurance, logistics, procurement
Counterparty reporting that today arrives as email attachments and gets retyped.
05 For the people who send you data
Your suppliers keep their format and get a straight answer
A supplier logs in, uploads the file they already produce and sees within minutes whether it was accepted. If something is wrong, the message names the field and the problem, in their language. Setup happens on your side, once per format. Submission is free for the sender.
Delivery · October 2026blocked
column "Net amount" · 41 of 212 rows
The values look like annual totals, about 12× the monthly amounts we expect for this column.
What to do: if these are annual figures, resubmit with monthly amounts; if they are monthly, confirm and we will accept the delivery.
Upload corrected fileConfirm values
06 Fits what you already run
An intake layer in front of your systems
DataOnboard replaces the manual consolidation step. Your ERP, warehouse and reporting models stay the system of record. Accepted deliveries leave as CSV, Excel or SAP export, through the REST API, through webhooks on every status change, or through connectors for PostgreSQL and Salesforce. AI agents can operate the portal through an MCP server.
Hosted in the EU.
Export
CSVExcelSAP export
Integrate
REST API v1WebhooksPostgreSQLSalesforceMCP server
Residency
EU-hosted
07 Pricing
Free to start. Pay in credits per accepted delivery.
Every month your organisation receives a set of free credits. Create a project, define a schema, invite a counterparty and run real deliveries on them. Each accepted delivery costs credits, depending on how many files it contains, their types and the processing they need. Suppliers never pay.
Free to start. No credit card.
08 FAQ
Questions we hear first
Will our counterparties actually use a portal?
They upload the same file they send you today and get a clear answer instead of silence. Nothing on their side changes, and submission is free.
Do you use AI to read the files?
AI helps when a new format is set up, by proposing the contract and the mapping. A person confirms it. After that, every delivery runs as deterministic SQL, so the result is reproducible and explainable to an auditor.
What happens when a file cannot be mapped to our schema?
The sender receives a field-level verdict describing what is missing or wrong. Your team sees the delivery as blocked with the same explanation.
We already have an ERP and a data warehouse.
Keep them. DataOnboard sits in front of them and replaces the step where someone merges files by hand. Clean, approved data leaves through your target format, the API or a connector.
Where is our data hosted?
In the EU. Every organisation's data is strictly scoped to that organisation; no other tenant can read it.
Do you handle PDFs?
Structured files (Excel, CSV, industry exports) today. Semi-structured documents such as PDFs are on the roadmap, with the source location recorded for every extracted value.
09
Run your first delivery today
Create a schema, invite one counterparty and see what comes back.
Free to start. No credit card.