Imagine running a reception desk at an international airport. Every passenger arrives speaking a different dialect. Some hand you a typed itinerary in plain English. Some show up with a handwritten note in a regional script you've never seen. Some bring a video, a verbal explanation, and a sigh. You're fluent in the airport's official customs language — that's your job — but every passenger has to be translated into it before they can clear the gate.
Now imagine the airport rotates its staff every two years.
That is what running a reinsurance ops desk looks like today — on SICS, on Sapiens, on Sequel, on every cession platform we've ever seen. SICS is the most common, and what follows is what the work looks like there.
DXC's Strategic Information and Cession System runs the reinsurance back office at over 100 reinsurers in 38 countries. It is a powerful, mature platform. It does many things well. But there is one thing it cannot do, and never has been able to do, and is not on a roadmap to do: read the unstructured documents your cedants actually send you.
Your cedants send you bordereaux as PDFs. Excel files with tabs that don't match. Emails with the bordereaux pasted inline. Sometimes the same data three different ways across two different files because two different teams own two different reports. Your operators read all of those and translate them into SICS entries by hand.
And here is the part that's harder to see from the outside: every operator on that desk builds a private dictionary in their head. Cedant A always reports premium net of brokerage in column F. Cedant B's profit commission true-ups arrive in narrative form, two months late, every Q3. Cedant C's claim bordereaux understate case reserves until the auditor reconciles them. That dictionary is institutional knowledge. And when the operator leaves, it leaves with them.
That is the real cost of the translator desk. Here are six entries that eat the day, and what the work to translate them really looks like.
The flagship line. Easy in theory, ugly in practice.
In SICS, written premium is its own entry — a clean number associated with a treaty period, currency, and cedant. In the cedant's bordereaux, that same number can appear three different ways: gross, net of taxes, net of commissions, with or without retro adjustments, sometimes split across multiple line items because the cedant's own system stores it that way.
Your operator's job: figure out which of those numbers should land in the written-premium entry, in which currency, against which treaty. Get one wrong and the cession accounting won't reconcile next quarter. The cedant who reports it that way isn't going to change. So the operator either knows — or relearns by making the mistake.
The "obvious" line that hides three things.
Commission looks straightforward — a percentage of premium, recorded as its own entry. But what counts as commission varies by treaty: ceding commission, override commission, profit commission earned vs. paid, and sliding-scale recalculations that change historical balances. The cedant may report all of these as "commission" in one column. SICS expects them broken out.
Your operator's job: split one column into three or four entries, and remember which treaty has which structure. The "remember" is the load-bearing word. With 30 cedants and 100 contracts, that memory is what separates a clean close from a forensic one.
The line that gets misclassified into commission.
Brokerage is paid to the intermediary, not the cedant. It is a separate entry in SICS — and a separate cash movement in your trust account. But many cedants combine commission and brokerage in their bordereaux because their own system doesn't distinguish.
Your operator's job: pull brokerage out of the commission column manually, every time, by reading the broker name and the contract structure. The operator who knows that Broker X has been on the contract since 2019 and their cut is always 2.5% gets it right in seconds. The new hire learns by getting it wrong.
The entry where errors compound silently.
Paid losses arrive in claim bordereaux — sometimes the same day as the premium bordereaux, sometimes weeks later. They have to be matched to the right treaty, the right loss event, the right cedant retention layer. Mis-bucketed paid losses inflate or deflate your loss ratios on the wrong contract. They show up not in the next reconciliation but in the next renewal pricing meeting, when the loss ratio looks wrong and nobody can explain why.
Your operator's job: read every claim line, decide which treaty layer it belongs to, and post it as a SICS entry. Multiply by every cedant, every month.
The entry that changes every month.
Case reserves are not a transaction; they are a snapshot. They go up and down as the cedant re-estimates. Each month's bordereaux contains a new estimate, and the SICS entry is the delta from last month, not the absolute number.
Your operator's job: pull this month's reserve, find last month's reserve, calculate the delta, and post it. If the cedant restates a prior month, redo the math for the prior period too. The operator who has done this for a particular cedant for two years knows the cedant's pattern of late restatements; the operator who is new to that cedant misses one in three.
The entry every operator dreads.
Premium adjustments are the catch-all: deposit premium reconciliations, MDP true-ups, retro adjustments, audit adjustments, swing-rated recalculations. They arrive late, they arrive in narrative form ("attached please find Q3 audit adjustments"), and they often span multiple prior periods.
Your operator's job: read the cedant's narrative, decide what's actually being adjusted and when, and post the right entries against the right historical periods. There is no shortcut. There is only the dictionary the operator has built up — or hasn't.
That is the desk job. Six categories of entry, multiplied by every cedant, multiplied by every reporting period. A reinsurer with 50 cedants is processing 300 entry decisions a month at minimum — and most reinsurers have more cedants than that, more entries than six, and more frequent reporting than monthly.
Every one of those decisions is made by a human, on a screen, after reading a PDF or an Excel tab. Every one of those decisions is correct in proportion to how well the operator remembers each cedant's quirks.
When the operator leaves, the cession-accounting team's accuracy goes down — not because the new hire is worse, but because the dictionary has to be rebuilt one mistake at a time.
Your operators are translators. They are taking documents written in a hundred different cedant dialects and rewriting them into the one official language SICS understands.
A few things follow from that:
A good translator gets better with every document. A working memory of each cedant accumulates — slowly, painfully — over months and years.
This is the value that's invisible from a job description but obvious to a CFO who has watched their senior operator retire.
The Brisc Reconciliation Analyst is that translator. But unlike the human version, the Reconciliation Analyst's working memory stays in the system.
This is why we describe the Brisc Reconciliation Analyst not as a productivity tool but as institutional memory for the cession-accounting desk. The output is correct entries. The asset — the part you carry forward — is the accumulating dictionary.
If you've been following ACORD's ADEPT initiative, you know the industry is investing in structured digital exchange between brokers, cedants, and reinsurers — and DXC has integrated ADEPT into SICS. Digital reinsurance transactions on ADEPT grew 73% from 2024 to 2025.
ADEPT is real, and it matters. But it solves a different problem. ADEPT works when both sides — cedant and reinsurer — have agreed to exchange data in a structured ACORD-compliant format. That is roughly 20% of the market today, growing.
The other 80% still send PDFs, Excel files, and emails. They are not on ADEPT. Many of them won't be on ADEPT for years.
ADEPT cleans up the structured 20%. The translator handles the unstructured 80%. They are complementary. A reinsurer should be doing both — and the dictionary the translator builds is itself the foundation that makes an eventual ADEPT migration cleaner. The cedant quirks Brisc has learned become migration logic when those cedants finally onboard to ADEPT.
Reinsurers who replace their translator desk with a system that retains its memory typically see, in roughly this order of importance:
The faster bordereaux turn-around is real. But it is the consequence of getting the four points above right — not the value itself. A reinsurer that buys speed without accuracy and without retained knowledge has bought the wrong thing.
We've seen deployments live in 2-6 weeks depending on cedant complexity and the dictionary's coverage at start.
Brisc is not a SICS replacement. It does not compete with DXC. It is the layer between cedant documents and SICS — the translator that turns the dialects into the official language. SICS is the system of record. It stays.
Brisc is also not a generic AI tool, and not a productivity tool. It is purpose-built for reinsurance — for bordereaux, for entry-code classification, for the specific reconciliation logic that makes cession accounting work. A general LLM cannot do this; a horizontal RPA tool cannot do this; a custom internal build will get the dictionary back to where your senior operator already had it but no further. The asset that compounds is the reinsurance-native dictionary.
Q: Does Brisc need access to our SICS instance? No. Brisc reads cedant documents and outputs in SICS upload format. Your team imports those uploads using the same workflow you use today. No SICS integration required for the core deployment.
Q: How does it learn our specific entry codes? Each reinsurer has a per-customer dictionary — a configuration of how Brisc maps fields to your specific SICS entry-code structure. We build it during onboarding (typically 2-6 weeks), refine it during the first few cedant runs, and it improves as new edge cases get reviewed. The dictionary is yours; it persists across staff changes; it compounds with every new cedant.
Q: What about cedants we haven't seen before? Brisc is trained on bordereaux as a category, not on individual cedants. New cedants generally process out of the box; the unusual ones get flagged for review and added to the dictionary on first pass. Every new cedant your team onboards joins the dictionary; the next operator who joins your team starts with that knowledge already in place.
Q: How do we know it's right? And what's the audit trail? Every entry classification ships with a confidence score and a complete source-document reference. Low-confidence entries route to your operator for review; high-confidence entries process automatically. The audit trail captures the source document, the extracted fields, the classification logic, the SICS entry that resulted, and any operator review applied. Regulators, auditors, and your own quarterly reviews can reconstruct any decision after the fact. Confidence-scored, source-traceable, fully reproducible — by design.
Q: Will this let us write more business without hiring? That is the structural advantage. Brisc's accumulated dictionary means each new cedant, each new program, each new geography absorbs into your existing operations rather than triggering an operations hire. One named Brisc customer (Helix Underwriting Partners) reports an 80% reduction in manual labour on submissions intake; the math at the cession desk is similar. The book grows; the team doesn't have to.
Q: Is this only for SICS? No. The architecture is the same for any reinsurance admin platform — Sapiens ReinsuranceMaster, Verisk Sequel, in-house systems. The dictionary changes per platform; the translator doesn't. SICS happens to be the largest deployed platform, which is why this post focuses on it.
If you are a reinsurer running SICS, your operator team has spent today translating. They will spend tomorrow doing the same. The bottleneck is not their effort; it is the model — humans translating dialects one document at a time, and carrying the dictionary in their heads when they leave.
The dictionary should belong to the reinsurer, not to whoever happens to be at the desk this quarter.