B BROCENT
All Services

BCS Beam · Self-Hosted · Private Deployment

Remote support, on infrastructure you control

BCS Beam is the remote support platform Brocent builds and operates itself — one unified device registry, two connection engines split by scenario: everyday assistance runs on in-person consent, unattended access runs on strong identity verification. Built to replace traditional subscription remote-support SaaS with a stack customers can audit, self-host, and license per seat instead of per subscription.

TL;DR: BCS Beam is Brocent's own remote support platform, delivered with Brocent Managed IT — session data never leaves the region the customer designates. This page walks through the full capability set: the dual-engine architecture (in-person consent vs. strong-identity unattended), four independent authorization channels, device binding plus a per-connection one-time code, policy-driven session recording and audit trails, a three-region self-hosted relay network, UI previews, and the complete lifecycle of both connection types.

156

Endpoints running the resident agent

4

Unattended authorization channels

3

Self-hosted relay regions (HK/SG/CN)

100%

Connections need consent or a per-session code

Why not a traditional subscription remote-support SaaS

Structural differences, not a feature-list difference

Matching features line by line is only the starting point. The real differences are: who owns the relay infrastructure, who can read session data, and what the cost model of every single connection looks like.

Data sovereignty

Self-hosted relays — no third-party servers in the path

Identity model

Zero trust: device binding + a per-connection one-time code

Pricing model

Licensed per seat — no year-on-year subscription creep

Internal-system integration

Direct links to tickets / contracts / customer master data

Brand & presentation

Fully white-labeled — customers only ever see their own brand

Overall architecture

One device registry, two connection engines split by scenario

This is not two unrelated tools glued together — every endpoint shares a single device identity across both engines from first install, and the system decides which authorization logic applies based on whether the device owner is present.

Both paths converge into one connection audit trail and one recording policy — the customer sees a single, coherent audit record, not two logs that don't reconcile.
🖥️

Daily Assist Engine

Scenario: routine support with the owner present

The resident agent deploys silently at install; the support team connects straight from the browser to view and operate — built for always-reachable, fleet-visible daily operations. Before control is established, the endpoint itself must show a confirmation prompt that the user explicitly allows.

  • · Direct browser connection — no dedicated client needed
  • · Concurrent multi-viewer sessions with handover
  • · Auto-grouped by customer, live device-tree status
  • · One-time share links for account-free ad-hoc collaboration
🔐

Strong-Identity Unattended Engine

Scenario: owner absent / high-compliance access

Built for support that must work with nobody at the keyboard — after a one-time hardware binding, every single connection still requires its own one-time dynamic code and must clear one of four authorization channels.

  • · Four independent authorization channels, from approval flows to admin direct-connect
  • · Policy-matched sessions are force-recorded; playback goes through a separately authenticated proxy
  • · Access windows + IP allow-lists — out-of-window connections are refused outright
  • · View-only observer mode that disables all control capabilities

Core capabilities

More than screen sharing — every connection is decided, filtered, and logged

Each capability below is tagged with the engine it belongs to:

Strong-Identity Engine Daily Assist Engine Shared by both
🔐

Device trusteeship

Personal and company-asset devices get a one-time, hardware-level binding for unattended access — yet every connection still requires a per-session code.

Strong-Identity Engine
🛡️

Four independent authorization channels

Engineer request-and-approve, email fallback verification, admin direct-connect, and view-only mode — none substitutes for another; each covers a different real-world scenario.

Strong-Identity Engine

In-person user consent

Daily assistance assumes someone is present: the endpoint shows a confirmation prompt before control is established, and a refusal terminates the connection immediately.

Daily Assist Engine
⏺️

Forced session recording

Sessions matching a customer's compliance policy are force-recorded and stream-archived, auto-purged at retention expiry, with playback through a separately authenticated proxy.

Strong-Identity Engine
📼

Server-side session recording

Daily-assist sessions can also be recorded server-side on demand — independent of whether the engineer's own machine has recording on.

Daily Assist Engine
⏱️

Access windows & IP policy

Connection windows set per customer time zone plus IP allow-lists — out-of-window attempts are refused before a session ever forms.

Strong-Identity Engine
👁️

View-only / observer mode

Blocks keyboard, mouse, clipboard, file transfer, audio capture, and remote reboot — while the target machine shows a system notification.

Strong-Identity Engine
🗂️

Grouping by customer

Devices are grouped automatically on enrollment — support sees a structured customer → device hierarchy, not a flat list.

Daily Assist Engine
🧭

Cross-engine device recognition

The paired strong-identity client ID on the same endpoint is auto-detected, so raising a high-compliance session never involves copying IDs by hand.

Shared by both
👥

Multi-viewer / session handover

Multiple engineers can view and collaborate on the same target device concurrently, with control handover that never drops the connection.

Shared by both
🧩

Install-free ad-hoc support

An install-free applet for one-off support scenarios — the backend never touches the real connection secret.

Strong-Identity Engine
📤

Offboarding review queue

Devices are never silently purged when a contract ends — they enter a human review queue, and agent removal only executes after confirmation.

Daily Assist Engine
🎫

Seat license management

Seats allocated per device, per user, or by concurrency — usage computed in real time, exportable to CSV.

Strong-Identity Engine

Identity & unattended access

First question: who is present? Two completely different verification paths

Unattended does not mean weakly verified — device binding answers "is this an authorized machine", and the per-connection dynamic code answers "is this specific connection an authorized person". When someone is present, the verifying party simply becomes the end user themselves.

Daily assistance: the end user decides, in person

Denial means termination — no gray zone

  • There is no “connect first, ask later” intermediate state
  • With nobody present, the flow switches to the Strong-Identity Engine — it never bypasses on-device consent
  • Every Allow / Deny choice is itself written to the operations log

Strong-identity unattended: four channels for four real-world scenarios

Channel 01

Engineer request & approval

With the owner away, the engineer files an access request; the system emails the device's registered contact, who can approve via link or release the session with a 6-digit confirmation code.

RequestEmail to ownerApprove / codeConnect

Channel 02

Email fallback verification

When a staff member's phone authenticator is unavailable, the system issues a one-time email code as the substitute credential.

Fallback requestEmail codeConnect

Channel 03

Admin direct-connect

A customer-portal administrator may connect without user confirmation only where the contract explicitly authorizes it — gated by admin role + MFA + a per-session code, with two-way notification afterwards.

Contract flagRole + MFAOne-time codeConnect

Channel 04

View-only observer

Once the session is up, keyboard / clipboard / files / audio / remote reboot are all force-disabled, and the target machine shows a system-level notification.

ConnectTarget notifiedControls disabled
Device-level binding answers “is this an authorized machine”; the per-connection dynamic code answers “is this specific connection an authorized person” — both checks must pass before a connection is established.

Recording & audit compliance

Recording is the outcome of a policy match — not a switch anyone can flip off

Each engine's recording mechanism writes into the same traceable session record — the customer receives one coherent audit thread, not two logs that don't reconcile.

📼

Forced recording (Strong-Identity Engine)

Policy-matched sessions are stream-recorded end to end into object storage — no reliance on local files, and the initiating party cannot switch it off mid-session.

🎬

Server-side recording (Daily Assist Engine)

Recording happens on the server, independent of the engineer's machine — no “forgot to record” and no “turned it off halfway”.

🗓️

Retention with auto-purge

Recordings and logs are purged automatically on the retention schedule agreed with each customer — audit-ready without unbounded archives.

🔎

Separately authenticated playback

Playback goes through its own proxy and permission check — viewing a recording is itself a logged access event.

🧾

Operation event log

Connect, disconnect, consent / denial, file transfer, and other key events carry their own timestamps, searchable independently of the recording.

📋

Genuine connection evidence

Every connection's start/end time, duration, verification method, and view-only status are recorded — not weak evidence like “a link was sent once”.

Global relay network

Three self-hosted relay regions — connections never touch third-party servers

The relay keys are held and signed by Brocent itself — this is not a public relay. Screen content and file transfers from both engines only ever traverse infrastructure the customer has approved.

Mainland-China cloud egress restricts overseas UDP traffic by default, so the mainland node connects back to Hong Kong over a TCP-carried encrypted tunnel — verified end to end.
🌐

Why not a public relay

Whose servers your screen content and files pass through decides who holds data sovereignty. Running our own relay services means:

  • · Relay keys are self-held — no third-party server can decrypt any relayed content
  • · Online-presence state uses shared multi-node storage, so multiple nodes are genuinely live at once
  • · A self-hosted update-check service — client updates never depend on a third-party distribution channel

Device grouping & operations

Organized by customer — not one flat device list

Devices joining the Daily Assist Engine are grouped under their customer immediately; a contract ending never silently purges devices — they enter a human review queue, and agent removal only executes after confirmation, so in-use devices are never deleted by mistake.

Device groups (sample data)

Customer groupDeviceStatusLast connected
Sample Group A FIN-WKS-0412 Online 2 min ago
Sample Group A FIN-WKS-0413 Offline 3 days ago
Sample Group B SRV-EDGE-021 Online Just now
Sample Group C POS-STORE-118 Online 16 min ago

UI preview

One brand and visual language, from client to admin console

The mockups below use sample data to illustrate the information structure the product presents — they are not screenshots of a real customer environment.

BCS Beam — strong-identity client
BCS BEAM v1.3.9-beam

This device ID

823 456 719

One-time password ready — awaiting remote connection request

Enter a remote ID to connect
Connect
End-to-end encryptedSelf-hosted relay
Strong-identity client home — real brand colors and logo Illustrative mockup
Remote session — observer mode
Connected by Engineer · A. Zhang Recording
Screen content (illustrative)
View KeyboardMouseClipboardFile transferRemote reboot
Remote session toolbar Grayed items are disabled in observer mode
remote-ops.internal — Daily Assist console
Sample Group A
FIN-WKS-0412
FIN-WKS-0413
Sample Group B
SRV-EDGE-021
Sample Group C
POS-STORE-118
Remote screen (illustrative)
desktopfilesterminalshare-link
Device tree + remote-view console Illustrative mockup
On-device confirmation prompt
🖥️ Remote assistance request

Support engineer L. Wei requests to view and control this device's screen for ticket #TCK-58231. Allow?

Deny Allow
End-user consent dialog Illustrative mockup
BCS admin console — seat licenses

Seat license allocation

+ New allocation
CustomerLicense typeUsed / totalStatus
Sample Group A Per device 62 / 100 Healthy
Sample Group B Concurrency 44 / 50 Near limit
Sample Group C Per user 18 / 60 Healthy
Seat license admin (sample data) Per device / per user / concurrency — three allocation bases

The full lifecycle of both connection types

From request to archived audit trail

The steps below cycle in real processing order (an animated illustration in place of a screen recording).

Daily Assist Engine Routine support with the owner present

01

Select device

Pick an online endpoint under its customer group

02

Initiate

Engineer clicks Connect to raise the request

03

Endpoint prompt

The device shows an Allow / Deny confirmation

04

User consents

Clicking Allow releases remote control

05

Session runs

Screen control, file transfer, recording where required

06

Disconnect & archive

Connection ends; the event log is written to the trail

Strong-Identity Engine Unattended / high-compliance access

01

Raise request

Engineer initiates the remote request from a ticket

02

Verify identity

Obtain the per-session code / email fallback code

03

Authorize

Access window + IP allow-list + channel permission checks

04

Connect

Session established via the self-hosted relay nodes

05

Policy applies

A recording-policy match starts streaming capture

06

Close & archive

Audit trail written; recording stored until retention expiry

Put remote support on infrastructure you can audit

BCS Beam is delivered with Brocent Managed IT — endpoint rollout, authorization policy, and recording retention are all configured to your compliance requirements. Contact us for a demo or a scoping conversation.

← Back to the BCS Beam Endpoint Support overview

All interfaces and data on this page are illustrative, anonymized samples — no real customer information is shown.