Ask across every product in your own words

Beryl Context searches your contacts, conversations, files, access records, tables, and invoices for one question, then links each part of its answer to the record that it comes from.

Search every product with one question

Beryl Context searches People, Channels, OneCloud, Teams, Tables, and Wallet for every question that you ask. When you ask for everything about a customer, a project, or a partner, Beryl returns their profile, threads, files, and access history together in one answer. Each result links to its source record, so you can open the source of any claim before you repeat it to a customer.

Ask once and open every matching record

The steps below show how Beryl Context answers a question, groups the results, links each one to its source, and narrows them with filters.

What Beryl Context finds across your workspace

Questions in plain language across every product illustration

A keyword search returns only the records that contain the exact words that you type. When a document has a different title, a thread uses a nickname, or a contact changes their email address, a keyword search will leave that record out of its results.

When you type a question the way that you ask a colleague, Beryl searches each product that your account can open for the records that match what you mean. You can ask for every contact named in a Q2 renewal thread, every file that anyone on your team shares with a partner over the past month, or every colleague who opens a given folder this week.

Your records about one customer, such as Halden Freight, can include their primary contact in People, six threads from the last quarter in Channels, the signed contract in OneCloud, and a shared workspace in Teams. To see all of it together, you will have to open each product in turn.

One query returns contacts, conversations, files, access records, table entries, and invoice history in 10 seconds. For Halden Freight, Beryl shows the primary contact beside the last thread, the most recent file, the pipeline record from Tables, and the latest invoice from Wallet.

When a search tool indexes every record and returns matches without checking who is asking, a junior analyst can open an executive memo through it, while a contractor can read a customer's personal data. You will usually learn about such an exposure only when someone reports it.

Beryl checks your access rules in Teams on every query and includes only the files and contacts that you can open directly. When the sales team keeps a contact private, Beryl shows that contact in search results to the sales team alone.

Filter by date, location, and type illustration

In a workspace with 4,000 documents, 18,000 messages, and 900 contacts, a keyword search can return 300 results, while all of them can come from outside the quarter that you need.

You can narrow a query to the last 30 days, the Lagos office, PDF files, or contacts tagged as partners. Beryl applies those filters as it searches each product, so the answer includes only the records that match both your question and your filters.

A search that returns a flat list can mix a dozen threads, a dozen files, and several contacts together, so you will have to work out which result belongs to which customer.

Beryl groups the results under the people and organizations that they concern. A question about late payments returns the five customers involved, each with their thread, their invoice file, and the colleague assigned to them.

What changes when every answer has a source

Every record about a customer in one answer

Beryl searches every product for one question and returns a single answer that you can check, copy, or share, so a single search finds the records, whichever products store them.

Answers with their sources attached

Each citation links to the file, thread, or profile behind the answer, so you can show anyone who asks where a claim comes from.

One set of roles for search and for every product

Beryl applies the access rules that you already maintain in Teams to every query, so search needs no separate set of permissions.

Results grouped by customer and partner

Because Beryl sorts each set of results by customer or partner, you can read them one relationship at a time.

How teams can use Beryl Context

Ten minutes before a renewal call, an account executive needs the whole history

Ten minutes before the call, the account executive asks Beryl for everything about the customer from the last 90 days. Beryl returns the primary contact and their role changes, three threads about pricing, the most recent proposal, two support tickets from Channels, and the colleagues who work on the account, with a link from each result to its source.

Result

The account executive starts the call with the pricing objection, the feature request from the champion's last meeting, and the most recently closed support thread on one screen. Because Beryl gathers those records from one question, the account executive prepares for the call from that single answer.

Ten minutes before a renewal call, an account executive needs the whole history

A support rep finds the answer that a former colleague gave two years ago

A support rep receives the same question about a feature that a former colleague answered two years ago. Beryl searches the threads in Channels, the help articles in OneCloud, and the earlier answers to the same question, then returns that colleague's resolution with a citation to the thread that contains it.

Result

The support rep replies with the resolution that the team already uses, along with a citation to the original thread, so they can skip asking around the team for it. New reps can find answers from former colleagues as well, because Beryl searches every thread that Channels stores.

A support rep finds the answer that a former colleague gave two years ago

A discovery request covers 14 months of records about one supplier

Opposing counsel asks for every document and message about one supplier over a 14-month period. Beryl searches the files in OneCloud, the threads in Channels, and the contact profiles linked to the supplier. The legal team exports the results with source citations, timestamps, and the access roles that applied when someone created each record.

Result

The paralegals collect the discovery records from a single query. Each exported record includes its source citation, so the legal team can show the court where every record comes from.

A discovery request covers 14 months of records about one supplier

A hospital billing analyst works out why an insurer denies a claim

When an insurer denies a high-value claim, the revenue-cycle analyst has to work out why. Beryl retrieves the patient account from People, the authorization and appeal correspondence with the insurer, the coding documents and prior-authorization letters from OneCloud, and the names of the billing staff who work on the account. Beryl masks personal health data in any answer that the analyst shares outside the billing team.

Result

The analyst builds the appeal packet from one answer, which lists the correspondence, the coding documents, and the authorization letters with their sources. With those records in one place, the billing team can resubmit a denied claim before the insurer's appeal deadline passes.

A hospital billing analyst works out why an insurer denies a claim

Ask Beryl Context about any customer

Ask for early access to search every Delalify product from one box, with a link to the source of every result.