Line by line
Every statement line matched to an invoice by hand, every period.
Partial and combined payments
One transfer for three invoices, or one invoice paid in two parts, and nobody wrote down which.
References that do not match
Customers type their own reference, the bank shortens it, the ERP expects another.
Several ERPs, several banks
Each market reconciles in its own system, and the group view waits for all of them.
Open items that stay open
Unmatched payments pile up and reminders go out for invoices that were paid.
No trail
When auditors ask why a payment was booked against an invoice, the answer lives in someone's head.
Bank reconciliation
Bank reconciliation, matched by AI. Your team confirms.
The bank statement comes in and the model proposes which open invoice each payment settles, in Business Central, Dynamics 365, legacy ERPs and bank exports. Your accounting team confirms matches instead of hunting for them.
Our team replies within two business days. Scope and price agreed before we start.
Tell us what you need.
Where it slows down
Reconciliation eats the days you need for the close.
Proven in production
Tested on real statements, not a demo file.
- less manual entry at Bredent Medical, across four markets
- 60%less manual entry at Bredent Medical, across four markets
- European markets on Dynamics 365 and legacy systems, one view
- 4European markets on Dynamics 365 and legacy systems, one view
- match rate on payment reconciliation at a betting and gaming operator
- 99.25%match rate on payment reconciliation at a betting and gaming operator
- transactions a month in that same reconciliation
- 4.6Mtransactions a month in that same reconciliation
Bredent Medical is named with permission. The match rate comes from a separate production system for an operator we do not name.
What it looks like
Bank reconciliation in 30 seconds
Silent, with captions. What the agent reads, what the rules decide and where a person steps in.
How we start
One bank, one entity, real statements. Then the rest.
01
Pick the account
One bank account and one entity with volume and a person who owns the reconciliation.
02
Run it on your statements
The model proposes matches on last month's real statements, so you see the hit rate before anything goes live.
03
Connect it
To your ERP and bank exports. Confirmation stays with your accounting team; nothing books without them.
04
Add accounts and markets
The next account goes faster than the first. Your team owns every rule.
Works with what your team already runs
- Dynamics 365
- SAP
- NetSuite
- Legacy and local ERPs
- Bank exports (MT940, CAMT, CSV)
- Payment providers
- Excel and CSV
Why SIEL
We built this for a distributor with four markets and four sets of habits.
- Your systems stay
- We connect Dynamics 365 and the legacy ERPs instead of migrating them.
- AI proposes, people confirm
- Every match is a proposal until someone in accounting confirms it.
- Traceable
- Each match links the statement line, the invoice, the rule or model output and the person who confirmed it.
Start the conversation
Which account takes the longest to reconcile?
Tell us the bank, the ERP and the monthly volume. Our team reads it and replies within two business days.