Behind every global investment portfolio is a mountain of transaction and reclaim data that needs to be gathered, cleaned, and matched before a single reclaim can be filed. It's essential work, and, until now, much of it happened out of view, in separate tools, before the results ever reached Protocol. With this release, that entire process now lives inside Protocol itself, and it's faster, more accurate, and far more transparent than before. 

Data cleaning and import, now inside Protocol 

Data cleaning and import used to sit outside the core platform. We've pulled both directly into Protocol, so the full journey of a data file - from upload to a matched reclaim position - happens in one system, without manual handoffs in between. The result is a process that runs significantly faster and produces cleaner, more complete matching from the outset. 

Reclaims matching is now AI-assisted 

Protocol's matching engine still starts by looking for exact matches between reclaims and income records. What's new is what happens next: where a reclaim doesn't line up exactly - say, the gross amount or income rate differs slightly from the income record - Protocol now uses AI to weigh the surrounding context and suggest a match, but only when the evidence points strongly in one direction and no other income record is a plausible alternative. The result is more reclaims matched automatically, without trading away accuracy. 

A full import log, for accountability 

Every data file that enters Protocol is now logged automatically: which file, imported by whom, and when. Clients and their teams can see exactly what has come in and when, giving them a clear, auditable trail of the data underlying their reclaim positions at any point in time.  

Disregarded and suspended records, out in the open 

Not every record in a data file belongs in a reclaim analysis, and not every record is immediately ready to be processed. Protocol now surfaces both categories explicitly: 

• Disregarded records — records that fall outside the scope of the reclaim analysis, such as currency transactions, income reversals, or the offsetting pairs of correction bookings. 

• Suspended records — records that are relevant to the analysis but need further clarification from the custodian before they can be processed. 

Rather than disappearing into the background, both are now visible within Protocol, giving clients full visibility into what's been excluded, what's on hold, and why. 

Knowing what's worth fixing first 

Not all data errors carry the same weight. Protocol now attaches an estimated recoverable WHT value to records with an error or omission. For example, 67 records where the net amount doesn't match gross amount minus withheld tax, worth an estimated EUR 475k in total. That number makes clear which issues deserve attention first. On the other end, a large batch of records missing a currency or ISIN might carry an aggregate estimated value of only a few thousand euros: a strong signal that those records can reasonably be suspended for later clarification, or disregarded altogether. 

What this means for clients 

Data cleaning, import and reclaims matching have always been part of the Protocol service: clients have never had to manage this process themselves, and that hasn't changed. What has changed is what clients experience as a result: the process is now considerably faster, more accurate and complete, and — thanks to AI-assisted matching, estimated recoverable value on data issues, the import log, and full visibility into disregarded and suspended records — more transparent than ever before.