Every submission, ready for a decision.
Email, broker portal, API or your own systems. Every submission is read, enriched, cleared and checked against appetite the moment it lands.
Underwriters start at the decision with the evidence attached, and spend the day quoting instead of keying.
Hours of intake, gone.
Submissions are sorted, read and matched to the account before anyone opens them. Your team spends the morning on quotes, not on the inbox.
No duplicate work.
A message you have already seen never opens a second submission, and junk is dropped before it is stored. Nobody triages the same thing twice.
Know what is missing on day one.
Every attachment is identified on arrival, so the queue shows what came in and what did not. Ask the broker for the gap while the account is still warm.
Renewals find their account.
A reference number, an existing thread or a known broker ties the submission to the account you already hold. Anything new opens as new business, never as a guess.
Nothing reaches the book unreviewed.
Extracted values arrive as a proposed change your team can read line by line. The book only moves when someone applies it.
“The mail used to be the job. Now it is sorted by the time we sit down, and what is left is the decision we are actually paid to make.”
Clearance without the detective work.
Each submission is matched to the right account on the strongest evidence available. When the evidence is split, an underwriter makes the call instead of the system guessing.
Matched by the number in the mail.
A policy or submission number quoted anywhere in the thread is checked against the accounts you hold first. The most reliable match is never second-guessed.
Every follow-up lands in the right place.
Once a thread belongs to an account, every later message joins it. Brokers reply however they like and the account history stays whole.
Known brokers, routed instantly.
The sender is matched against the brokers and insureds on your accounts. One match routes it. Two, and a person makes the decision.
A queue your team can actually clear.
See what is waiting on a person, filter to what matters, and fix any submission that was read wrong without starting over.
Find any submission fast
Narrow the inbound book by channel, status or date, and page through it without rows shifting under you.
Stuck work is never hidden
A submission that needs a person says so, kept apart from the ones still moving.
The right owner, automatically
A submission that needs attention opens a task for whoever owns the inbox it came through.
Fix one step, keep the rest
Re-run a submission from the step that went wrong. Everything before it is kept, so only the wrong part is redone.
Live, with no refresh
The queue updates as each step lands, so the team always works from the current picture.
Every decision is on the record
The candidates considered, the confidence and who resolved it are all kept, so any routing decision can be explained months later.
Every channel lands in one queue.
Take submissions however your brokers and systems send them. Where a submission came from is recorded, and it never changes how it is handled.
Keep the inbox brokers already use.
Connect the mailbox submissions already go to. Nothing changes for your brokers, and an inbox can be scoped to the programs it actually receives.
Give producers a front door.
Producers submit through a portal that asks for exactly what the program needs. Complete submissions come in, and fewer emails go back and forth.
Send submissions from your own systems.
Post submissions straight from a rater, an agency platform or your own tools. They join the same queue and get the same checks.
Bring in what lives elsewhere.
Use workflows to pull submissions from the systems your team and partners already run. Nothing waits on a manual export.
Decision-ready, not just data-entered.
Once the documents are read, the platform runs the checks your team runs by hand today. Underwriters start from the answer, and a risk you will never write never costs a quote.
Fuller accounts, no research.
Each submission is enriched with what your team would otherwise look up: the business, its locations and the parties around it. Underwriters start with the whole picture.
Clearance before anyone reads it.
The insured is checked against the accounts and submissions you already hold, so a duplicate or a conflict is caught on arrival, not after a quote goes out.
Knockouts, stopped early.
Your knockout rules run as soon as the data lands. A risk you will never write is flagged before it costs anyone an hour, with the requirement it raised attached.
Appetite on every submission.
Each submission is scored against the appetite you configured and lands in appetite, referred or out. The queue shows what is worth your team’s time first.
“What changed my mind was clicking a number and landing on the exact box it was read from.”
Numbers you can defend.
Every extracted value shows where it was read and how sure the model was, and what becomes a record is still a decision someone makes.
Settle a disputed figure in one click.
Every extracted value keeps the page it came off and the coordinates of the box it was read from. The viewer draws that box back over the original, so a disputed figure is one click from its source rather than a search through an attachment.
Review time goes where it matters.
Each field comes back with its own confidence, colour-coded in the fields panel next to the document. The weak ones are the ones a reviewer sees first, so attention goes where the model was least sure instead of down the page in order.
Conflicting documents never get lost.
When two files answer the same question differently, both candidates are kept. You see which value won, that it won on confidence, and what lost. Nothing is quietly overwritten by whatever arrived last.
Better accuracy, no risk to live work.
Attach the values you expect to a sample and run the whole pipeline against it. Every stage gets a pass or a fail, the run never touches your data, and the prompts you tune are versioned and released like the rest of your configuration.
New programs without a project.
A document Riska has never seen is configuration, not a project. Define the documents your programs receive and the fields they fill, and everything captured joins the Data Mesh the rest of the platform reads from.
Any document, not a fixed list
Name the documents your programs actually receive and write the criteria that tell them apart. Scope one to a single product or to everything you write.
Setup that drafts itself
Point the setup at a few samples and it returns a field-by-field mapping onto your objects, prompt included. Nothing is written until you apply it.
New fields become new risk factors
An extracted field lands on your own model, so appetite rules, rating inputs, dataviews and the assistant can all use it the moment it arrives.
What happens next.
Triage hands over a cleared submission with the record already populated. The rest of the platform picks it up from there.
Send us a week of submissions.
Bring a real submission inbox, attachments and all. We will run it through and show you the decisions waiting on the other side.
Usually 30 minutes, with an underwriter on the call.