Smart importer
LiveBring your book across in an afternoon.
Export from OpenGI, Applied Relay or a spreadsheet and hand the file over. Columns, dates, policy types and consent flags are worked out for you, and you review the mapping before a single record is written.
- opengi-export-2026-07.xlsx — 1,842 rows
- → Policy type · 7 codes mapped to 5
- INFERRED
- → Renewal date · DD/MM/YYYY
- INFERRED
- → Marketing consent · Y/N
- INFERRED
- Nothing written yet
- HELD
Whatever your system exports.
CSV, XLSX and XLS, with the real format read out of the file itself, so a mislabelled export still lands correctly. Exports from OpenGI and Applied Relay are recognised by their shape, so a file you have produced a hundred times is understood on the first upload.
The mapping is the work. We do the mapping.
Working out that a column called POL_TYP holds policy types, and that its seven codes mean five things, is the part that costs a day of somebody’s life. That inference happens for you, on the columns and on the values inside them.
Nothing is written until you approve it.
The proposed mapping is shown back to you before any record enters the book. You correct anything wrong and approve it. There is no version of this where an import silently writes a thousand rows into the wrong fields.
Dates and consent come across too.
Renewal dates are parsed into real dates, not left as strings, and marketing consent flags are carried into per-channel consent records. Those are the two things most likely to be dropped in a migration, and the two you are most likely to be asked to evidence later.
Questions
Smart importer, in detail.
- Does AI Broker connect directly to OpenGI or Applied Relay?
- No, and it is worth being precise about that. You export from your existing system and upload the file; there is no live connection or API link to either platform, and nothing is read out of them automatically. What we take off you is the mapping work. That is the part of a migration that actually costs a day.
- Which file formats can I import?
- CSV, XLSX and XLS. The format is read from the contents of the file, so a spreadsheet saved under the wrong name still imports correctly.
- What if my column names are unusual?
- That is the normal case. Column names and the values inside them are both inferred, so a column called POL_TYP holding a set of internal codes is mapped without anyone writing a lookup table by hand. You see the result and correct anything that is wrong.
- Can an import overwrite my existing book by mistake?
- The mapping is reviewed and approved before anything is written, so an import cannot commit on its own. Nothing enters the book until a person has looked at the proposed mapping and accepted it.
- Do consent flags and renewal dates survive the import?
- Yes. Renewal dates are parsed into structured dates and marketing consent is carried across into per-channel consent records, so the brokerage can evidence what a client agreed to, instead of reconstructing it from a spreadsheet two years later.
Next
Revenue you already hold
A daily read across the whole book surfacing cross-sell gaps, dormant clients, insurer concentration and renewals worth a call.
Every enquiry answered, day or night
Structured intake across every product line and every channel, running out of hours, with a broker takeover that freezes the AI mid-thread.
One claim, one file
Customer, insurer and assessor threads on the same claim gathered into one matter, across whichever channel each party uses.
A serious AI for serious work
See your whole book before your competition does.
Independent and mid-market brokerages in Ireland and the UK. Onboarded in a week. Trained on your knowledge, not the open web. A 30-minute call is the whole sales process.
A 30-minute call is the whole sales process. Or email support@aibroker.ie.