Deployment
Call center deployment: keep the phone system you already paid for
Most contact center platforms answer "what about our PBX?" with a migration plan. DialerBee runs three ways: we supply the voice, you keep your own PBX, or you bring your own carrier. All three are live in production today.
Quick answer
What does DialerBee deployment cover? DialerBee runs three ways: we supply the voice on our own SIP trunks, you keep your own PBX or SBC and we send outbound traffic to your gateway, or you bring your own carrier. All three are live in production today. You pay per minute only when we carry the voice; keeping your own PBX or your own carrier means a per-agent platform fee, and your telecom stays on your existing contract at your existing rates. Compliance runs before origination, so consent state, DNC suppression and calling-window rules are evaluated by the platform whichever pattern you pick, and you can move from one pattern to another without a cutover weekend.
A floor with a working PBX and negotiated carrier rates is not looking for a reason to throw both away. The usual answer from enterprise cloud contact-centre software is that the voice has to move to them — which turns a dialer decision into a telephony migration, a procurement cycle and a risk committee.
You do not have to move the voice to get the dialer.
DialerBee can sit on top of the infrastructure you already run. The campaign engine, pacing, compliance enforcement, agent desktop, supervisor tools and reporting are ours. The trunks, the carrier contract and the call path stay yours.
Three ways to connect
Pick the one that fits what you already own
The pattern is chosen per customer, and it is a property of your account rather than a different product. Everything above the trunk behaves the same way whichever you choose.
We supply the voice
Simplest to startWe own and operate the SIP trunks. You get a dialer that works on day one with no telephony of your own to configure.
Fits you if
- New to contact-centre software, with no telephony estate
- Moving off a legacy on-premise dialer and wanting voice managed for you
- Smaller floors — roughly 10 to 50 agents — that want simplicity
- MENA operations that want local carrier relationships handled locally
What it involves
- Default gateway configured when your tenant is created
- Numbers allocated from our inventory and registered with the regulator
- No SIP configuration required on your side
Keep your own PBX
The one CCaaS cannot offerYou keep the PBX and the trunks you already paid for. DialerBee becomes the campaign, pacing and compliance brain that sits on top of them, and the voice never leaves your infrastructure.
Fits you if
- 50–500 agents with a PBX investment you are not about to abandon
- Carrier contracts and per-minute rates you already negotiated
- A regulatory requirement to keep voice traffic on your own infrastructure
- System integrators bundling a dialer with their own voice platform
What it involves
- Asterisk, 3CX, Cisco CUCM, Avaya or any standards-compliant SBC
- TLS, TCP or UDP signalling, with TLS strongly recommended
- IP allow-list, SIP digest or registration — your choice of authentication
- Your codec order, your DTMF mode, your concurrency ceiling
- Your numbers, ownership verified by Letter of Authorisation or test call
Bring your own carrier
Vendor diversificationYou buy SIP trunking from whichever carrier you prefer and give us the trunk credentials. We originate through it on your behalf.
Fits you if
- You already hold a carrier account with rates you negotiated
- You operate where we do not have a direct carrier relationship
- You deliberately want the dialer and the voice from different vendors
What it involves
- Gateway points at your carrier's SIP endpoint
- Same authentication, codec and capacity controls as a tenant PBX
- We carry no voice traffic and add no per-minute margin
Your PBX, in detail
What connecting your own PBX actually involves
Not much, and none of it is bespoke. Your telecoms engineer will recognise every field.
Endpoint
The host or FQDN of your SBC and its SIP signalling port.
Transport
TLS, TCP or UDP. TLS is strongly recommended and is what we default to in guidance.
Authentication
IP allow-list against your source ranges, SIP digest credentials, or registration to your SBC — whichever your security team prefers.
Codecs
Your preferred order. Opus first by default, falling back through G.722 and G.711.
DTMF
RFC 2833, in-band or SIP INFO, to match what your platform already expects.
Capacity
A maximum concurrent-call ceiling and a calls-per-second limit, so pacing never pushes your SBC past what it is sized for.
Health checks
A SIP OPTIONS probe so gateway health is monitored rather than discovered during a campaign.
Failover
Each gateway can name its own failover gateway, so a trunk problem does not stop the floor.
Your numbers stay yours. When you own the trunk you also own the caller IDs. We verify that you are entitled to present them — by Letter of Authorisation or a test call — before any number can be used, because presenting a number you cannot prove you control is the fastest route to a carrier complaint.
Identical either way
The pattern changes the trunk, not the product
Compliance gating
Consent, DNC and calling-window enforcement run before a call is placed, whoever owns the trunk.
Recording
Captured at the platform, before the call egresses to your gateway — so you keep recordings even when you own the voice path.
Agent desktop
The same browser softphone, disposition flow and script surface for every agent.
Supervisor tools
Live wallboard, listen, whisper and barge behave identically across all three.
Reporting
One set of CDR, campaign, agent and compliance reports — the pattern does not fragment your numbers.
Tenant isolation
Row-level isolation and role-based access are properties of the platform, not of the trunk.
Not a one-way door
You can change pattern without a cutover weekend
PBX contracts expire and priorities change. Moving between patterns is a campaign-by-campaign migration, not a big-bang switch.
A second gateway is created alongside the one you are already using.
A test campaign runs against the new gateway while live traffic carries on untouched.
Campaigns move across once the new path is proven, at whatever pace you want.
The old gateway is deactivated only when nothing is running on it any more.
How the platform is hosted
Whichever pattern you pick, the platform runs the same way
The trunk is your choice. What sits above it is not something you have to size, patch or babysit.
Two media nodes
Media runs on two FreeSWITCH nodes rather than one, so a single machine is not the whole platform.
Agent failover
The agent desktop holds a WebSocket to a media node. If that node goes away, the agent's client fails over to the other one at reconnect.
Load-balanced WebRTC
Browser-based agent audio is load balanced across the nodes, so agents are spread rather than piled onto whichever node answered first.
Round-robin SIP
A SIP load balancer distributes calls round-robin across the media nodes, so carrier traffic does not depend on one node staying up.
BYOC on any pattern
Bring your own carrier and the gateway points at your carrier's SIP endpoint, with the same authentication, codec and capacity controls as a tenant PBX. We carry no voice traffic and add no per-minute margin.
Nothing for you to run
The media nodes, the load balancing and the failover are ours to operate on every pattern, including the ones where the trunk is yours.
The reconcilers, the readiness checks and the backup regime behind this are written up on the platform reliability page.
Questions
What buyers ask about deployment
Can I keep my existing PBX and still use DialerBee?
What exactly do I configure for a tenant-PBX deployment?
Who owns the phone numbers?
Do I pay per minute?
Can I start on one pattern and move to another?
Does compliance still work if the voice is not yours?
Tell us what you already run
Bring your PBX model, your carrier arrangement and your agent count, and we will tell you which pattern fits and what connecting it involves, before you talk to procurement.