Question 1

What does Klarefi do?

Accepted Answer

Klarefi is a decision-ready intake platform that sits on top of your existing systems. Your applicants complete a branded form and answer contextual follow-up questions in a chat interface. Behind the scenes, Klarefi extracts facts from submitted documents, verifies them against your internal tools and databases, and assembles a complete, cited case file for your team to review and decide.

Question 2

What is a processed intake?

Accepted Answer

A processed intake is one complete submission that has gone through the full pipeline: form submitted, documents received, facts extracted with source citations, gaps identified through chat follow-up, verifications run against your systems, and a case file assembled for your team to review. A single intake can contain multiple documents. You pay per submission, not per page.

Question 3

What is a "fact" in Klarefi?

Accepted Answer

A fact is a specific piece of information your workflow needs — a name, a date, a policy number, a coverage amount, a database match. You define facts when you build your workflow: the label, description, expected type (text, date, enum, address, number), and whether to verify it against another fact or an external system. When documents come in, Klarefi locates each fact in the source text, extracts it, and attaches the exact quote and page number as evidence.

Question 4

How does pricing work?

Accepted Answer

Pricing is based on your intake volume and workflow complexity. We align cost to the value of decision-ready cases, typically starting with a pilot that has measurable targets. Volume commitments are available for high-throughput operations.

Question 5

Who is Klarefi built for?

Accepted Answer

Process managers and operations teams in regulated industries — insurance, banking, legal, healthcare, government, and corporate enterprise. If your team spends hours reading documents, cross-checking details across systems, and chasing missing information before making a decision, Klarefi handles that triage work.

Question 6

How long until we see results?

Accepted Answer

The proof sprint runs 2 to 4 weeks on your actual documents. You get measurable accuracy and coverage numbers by the end. Production deployment typically follows within 1 to 2 weeks after that.

Question 7

What happens after I request a pilot?

Accepted Answer

We assess fit within one business day. If your use case aligns, we schedule a short call to scope the pilot and define target metrics. We configure your FORM, FACT, and VERIFY workflow and run it on your actual documents. You get measurable accuracy and coverage numbers within 2 to 4 weeks.

Question 8

Does this replace our existing systems?

Accepted Answer

No. Klarefi sits on top of your core systems and handles the intake layer. Documents come in through your existing channels. Structured outputs push to your backoffice, DMS, or CRM via API and webhooks.

Question 9

What happens when something needs correction?

Accepted Answer

Your team reviews extracted facts and corrects them directly in the case file. Klarefi does not decide autonomously — approvals stay with your people. Corrections are used to refine your workflow's fact configuration and examples. They do not train underlying models. This improves accuracy within your specific intake while keeping your data in your control.

Question 10

Where is our data stored?

Accepted Answer

EU data residency by default. We store only derived artifacts and evidence pointers, not your raw documents. No training on customer data — contractually guaranteed. Configurable retention policies, audit logs, and access control included.

Question 11

What types of documents can you process?

Accepted Answer

Any unstructured document: PDFs, scanned pages, emails, photos, attachments. One case might have 5 to 30 documents depending on complexity. Klarefi structures them all.

Question 12

Do we need to provide training data?

Accepted Answer

No. Klarefi works on day one without training data. Workflows are configured through FORM, FACT, and VERIFY — not trained. The system improves from corrections over time, but it does not need historical labeled data to start.

Question 13

What if it doesn't work for our specific workflow?

Accepted Answer

That is exactly why we start with a Proof Sprint. We test Klarefi on your actual documents. If we cannot reduce your manual triage time by 60% within 30 days, we continue to work with your team one-on-one until we hit that metric.

Frequently Asked Questions

Frequently Asked Questions

Frequently Asked Questions