Question 1

What does Rivr do?

Accepted Answer

Rivr connects to the storage where a matter's files live and processes every file: OCR for scanned pages, classification into construction document types, splitting of combined PDFs into their individual forms, extraction of key fields from change orders, pay applications, RFIs, invoices and contracts, and resolution of the companies and people named across them. You then search, filter, group documents into issues, view them on a timeline, and export to Excel or connect your AI tools. Every value links to the page it came from.

Question 2

Who is it for?

Accepted Answer

Construction claims and disputes consultancies, and surety claims teams. The people who use it day to day are the consultants doing the analysis: quantum, cost to complete, delay support, expert work. The people who bring it in are the practice leaders who answer for the work.

Question 3

What kinds of documents does it handle?

Accepted Answer

More than 80 construction document types, out of the box: change orders and change order requests, pay applications, RFIs, invoices, contracts, schedules, meeting minutes, daily reports, correspondence and the rest of a typical production, including scanned PDFs. The key fields are extracted from the documents that drive the money on a matter: change orders, pay applications, RFIs, invoices and contracts.

Question 4

How big a matter can it take?

Accepted Answer

Rivr is designed for productions that run to millions of documents. It is built for a matter, not a folder.

Question 5

Does it work for surety claims?

Accepted Answer

Yes. After a default, Rivr brings the pay applications, change orders and invoices into one place and extracts what was billed, paid and retained, by subcontractor, so the cost to complete starts from the record rather than from a stack of PDFs.

Question 6

Where do our documents go?

Accepted Answer

Your files stay in your storage. Rivr connects to Box, Egnyte or SharePoint with the access you grant and reads the files there. The text and fields Rivr extracts are stored in Rivr, encrypted, isolated to your organization and the matter they belong to.

Question 7

Is our data used to train models?

Accepted Answer

No. Your content is never used to train or improve models, ours or anyone else's. Model providers process requests under zero data retention terms.

Question 8

How is access controlled?

Accepted Answer

Access is granted per matter. Only the people you add to a matter can see it. Administrators can set granular permissions by group and review audit logs of who viewed, downloaded or exported what.

Question 9

Is Rivr SOC 2 compliant?

Accepted Answer

A SOC 2 Type II audit is in progress, with controls managed in Vanta. We will share the report when it is issued and the current status on request. Data is encrypted in transit and at rest, environments are separated, and access-control tests run on every release. We would rather tell you what is in progress than claim what is not yet done.

Question 10

Which AI models does Rivr use?

Accepted Answer

Rivr is model agnostic. It chooses the model for each task, from leading providers, based on accuracy, cost and speed, and keeps the option to switch. You are not tied to one provider, and no provider trains on your content.

Question 11

Can we delete a matter?

Accepted Answer

Yes. You can delete a matter at any time, and the extracted data goes with it. Your source files were never moved, so they stay exactly where they were.

Question 12

Can Rivr be used on matters where the client contract restricts AI?

Accepted Answer

That is your call and your client's, and we will help you make it with the facts: what Rivr stores, where it runs, and what it never does with the content. Many firms start on a closed matter where the question does not arise.

Question 13

How is Rivr priced?

Accepted Answer

Per matter, with unlimited seats. Pricing is tiered by the number of files in the matter, so a small surety file and a two-million-document dispute are not priced the same. There is no per-user charge, so the whole team can work in it. Because the cost sits with the matter, many practices pass it through to the client as a technology fee.

Question 14

Is there a free trial?

Accepted Answer

Not a self-serve one. We start with a thirty-minute demo. If it fits, we run Rivr on a closed matter of yours at no cost so you can judge it against work you already know.

Question 15

What does a pilot look like?

Accepted Answer

One or two matters, live or closed, about thirty days from kickoff. You connect the storage or share a folder, we configure and process, you work in Rivr, and we walk through the results together at the end. Success is measured in review time saved and whether your team wants it on the next matter.

Question 16

Do we have to change how we store documents?

Accepted Answer

No. Rivr reads from Box, Egnyte and SharePoint and writes back to them if you want organized folders and file names there. You can also upload files directly.

Question 17

Can we keep working in Excel?

Accepted Answer

Yes. Export any table, with the extracted fields and a manifest that traces every row to its source document. The timeline and the organized matter stay in Rivr; the numbers go wherever you need them.

Question 18

Can we use it with Copilot, ChatGPT or Claude?

Accepted Answer

Yes. Rivr connects to those tools through MCP, so your team can ask questions, draft and summarize against the organized matter in the assistant your firm already approved. Rivr does the reading and the ordering; your tools and your people do the rest.

Question 19

How accurate is the extraction?

Accepted Answer

Every extracted value shows the exact spot on the page it came from, so you can see for yourself. Values Rivr is less sure of are flagged for review. You can verify or correct any value, and the correction stays with the matter. We publish no accuracy percentage because the right amount of checking depends on your matter and your standards, and we would rather you see the source than trust a number.

Question 20

Does it understand our naming conventions?

Accepted Answer

Rivr uses a construction taxonomy out of the box and can be set to your firm's or your client's terms, so if they call it a COR, Rivr calls it a COR.

Question 21

Does Rivr do delay analysis or tell us who is responsible?

Accepted Answer

No. Rivr puts the record in order and shows you what happened and when. It does not reconstruct schedules, assign fault or percentages of responsibility, or write the opinion. That is your work, and it stays yours.

Question 22

Will it replace our consultants?

Accepted Answer

No. It replaces the weeks of sorting, renaming and retyping that come before the analysis. The judgment, the investigation and the report still belong to the person whose name goes on it.

Question 23

Is Rivr a document management system or an eDiscovery platform?

Accepted Answer

Neither. Your documents stay in the storage you already use. eDiscovery platforms find words for lawyers; Rivr organizes a construction production into documents, fields, parties and issues for the people doing the analysis.

Question 24

Who is behind Rivr?

Accepted Answer

Two founders: Matt Tiberii, who spent his career in construction and lived the document problem as a project manager, and Brian Bolze, an exited software founder who builds Rivr's product. Rivr is backed by Gutter Capital.

Question 25

How do we get started?

Accepted Answer

You will get thirty minutes with the founders, on a synthetic matter or your own.

Questions

What Rivr is

Security and your data

Pricing and pilots

Working with Rivr

What Rivr does not do

Company

See Rivr on a matter like yours.