Integrations
There are three doors into this network: a SIP registration, an IP-authenticated trunk, or an HTTPS request. Everything below is one of those three, described specifically.
How a system connects
Pick the door that matches what you run. Most accounts end up using two of the three.
SIP registration
A username and a password, refreshed on an interval. Works from behind NAT and from dynamic addresses, which is why desk phones, softphones, and small PBXs use it. Per-trunk credentials, per-trunk access control list.
IP authentication
You give us the source addresses your traffic comes from and we allowlist them at the edge. The right choice for softswitches, proxies, and anything carrying volume, because there is no registration to expire.
REST API and webhooks
Bearer-token requests over HTTPS to provision numbers, originate calls, send messages, and query CDRs, with signed webhook deliveries coming back for call progress, recordings, transcripts, and receipts.
Everything named on this page describes protocol-level compatibility. It is not a claim of a partnership, a certification, a reseller agreement, or a marketplace listing with any vendor mentioned.
Softswitches and PBXs
Keep the system you already run. It points at a trunk hostname and nothing else about it changes.
Asterisk
A PJSIP trunk with registration or IP authentication. A complete working pjsip.conf is in the documentation, including the DTMF and media settings that matter.
How it connectsFreeSWITCH
A Sofia gateway with register set true for credential auth or false for IP authentication. Same codec and DTMF guidance as Asterisk.
How it connectsFreePBX
A SIP trunk defined in the GUI over the same PJSIP settings, with inbound routes pointed at your extensions and outbound routes at the trunk.
How it connects3CX
Added as a SIP trunk provider entry, authenticating by registration or by IP, with your own inbound and outbound rules on top.
How it connectsKamailio
Dispatcher or uac entries against the trunk hostname. IP authentication is the better fit at proxy volumes because it removes registration refresh from the failure surface.
How it connectsGeneric SIP
Any RFC 3261 user agent: session border controllers, softphones, desk phones, ATAs, and appliances. If it registers or authenticates by IP, it connects.
How it connectsCollaboration
Microsoft Teams connects by Direct Routing, so PSTN calling arrives inside the client your users already have.
- Direct Routing, not a bolt-on client. Calls are placed and answered in Teams. Users do not learn a second application.
- Your numbers, ported or new. Existing DIDs port in and are assigned to Teams users; new ones are provisioned by API.
- Per license, per month. Connectivity is billed per enabled license, and the calls and messages themselves rate at the published per-minute and per-message rates.
- Mixed estates are normal. Teams for the office and a hosted PBX or SIP endpoints elsewhere, on one account and one invoice.
Contact center and dialer platforms
This is the category the network was built for. AgentTech Dialer runs its production insurance calling on these rails, and Solved Telephony owns AgentTech Dialer, so the routing primitives here were designed against real dialer traffic rather than imagined from a spec.
Platforms in this category usually take the trunk for media, the API for provisioning, and the routing engine for bidding, caps, and skills. White-label is the default: your customers see your product.
AgentTech Dialer
Runs its production insurance calling on these rails. Solved Telephony owns AgentTech Dialer, which is why the routing primitives on this network look like a dialer built them.
How it connectsPredictive and power dialers
IP-authenticated trunks sized on peak simultaneous calls rather than agent count. Velocity limits and destination allowlists sit in front of the traffic.
How it connectsContact center platforms
SIP trunks for the media path, plus the REST API for DID provisioning and the routing engine for queueing, skills, and overflow.
How it connectsCall tracking and attribution
Buy and release DIDs by API as campaigns come and go, then take call outcomes from CDR queries and recording and transcription webhooks.
How it connectsVoice broadcast and IVR
Trunks with spend caps, velocity limits, and destination allowlists applied, and multi-level IVR run under your own prompts and brand.
How it connectsHosted PBX resellers
Sub-accounts per end customer, each with its own numbers, trunks, and CDR rollup, settling to your parent account.
How it connectsCRM and vertical SaaS
No connector directory to wait on. If a system can send an authenticated HTTPS request or receive a JSON POST, it integrates.
Any CRM with an API
Push call outcomes, recordings, and transcripts into the record after the call by subscribing to call.completed and transcription.available.
How it connectsAgency management systems
Write CDR-derived activity against a policy or a client record, and pull recordings by signed URL rather than copying media between systems.
How it connectsVertical SaaS
Embed number provisioning, routing, and recording inside your own product so a customer never leaves it to configure telephony.
How it connectsScheduling and field service
Confirmations and reminders over SMS from the numbers already on the account, with consent state read before each send.
How it connectsCustom internal applications
Register an HTTPS endpoint, subscribe it to the events you care about, verify the signature, and return a 2xx quickly.
How it connectsWorkflow and automation tools
Anything that can send an authenticated HTTPS request and receive a JSON POST can drive and observe calls without a purpose-built connector.
How it connectsData and warehousing
If you resell these minutes you need records you can bill from and defend. They go where your analytics already live, on a schedule or event by event.
Amazon S3
Scheduled CDR export written to a bucket you own, hourly or daily, as complete files per window.
How it connectsGoogle Cloud Storage
The same scheduled export, written to a GCS bucket. Load on the export window rather than the file timestamp and the arithmetic stays correct.
How it connectsScheduled CDR export
Every leg with its direction, route, carrier, duration, attestation, sub-account, and the rate that actually applied.
How it connectsWebhook streams
Event-by-event delivery to an HTTPS endpoint when you want the warehouse loaded continuously instead of on a schedule.
How it connectsAI and agent runtimes
Machine listeners are less forgiving than people, so the media path is treated as a product. Bring your own stack, or use the managed agent at $0.10 per minute of talk time.
Speech-to-text engines
Uncompressed mu-law media and symmetric RTP give a recognizer the cleanest input available. Or use network transcription at $0.05 per minute and skip the engine.
How it connectsText-to-speech voices
Synthesized audio is carried on a media path tuned for turn-taking rather than for hold music, with jitter buffers set for speech.
How it connectsLLM providers
Your runtime holds the model relationship and drives the call over REST while we carry the media. The latency budget is a few hundred milliseconds end to end.
How it connectsMCP-capable agents
An agent that speaks the Model Context Protocol can treat our REST endpoints as tools: search a number, place a call, read a transcript, query CDRs.
How it connectsRates for everything referenced here are on the pricing page, and the connection details are in the documentation.
FAQs
Integration questions
Does naming a system here mean you have a partnership with its vendor?
No. Everything on this page describes protocol-level compatibility: the system speaks SIP, or it can call a REST endpoint, or it can receive a webhook. None of it is a claim of a partnership, a certification, a reseller agreement, or a marketplace listing with any vendor named.
Do I need a developer to connect?
Not for a trunk. Pointing Asterisk, FreeSWITCH, FreePBX, 3CX, or any RFC 3261 endpoint at us is configuration, and a working example is in the documentation. You need a developer when you want provisioning, routing, and reporting to happen inside your own product, which is what the REST API and webhooks are for.
Can my users see that you are underneath?
Only if you want them to. White-label is normal here: numbers are provisioned under your brand, IVR and queue prompts are yours, and CDRs come back split by your own tenant identifiers. Most platform customers run entirely without our name appearing.
Is there a charge for the API, webhooks, or CDR exports?
No. The API, webhook delivery, and scheduled CDR exports are part of the account rather than an integration tier. You pay published rates for the usage they generate: minutes, messages, numbers, recording, transcription, and agent talk time.
What does Microsoft Teams cost?
Teams voice connectivity starts at $4.50 per license per month and messaging at $2.00 per license per month, on top of the standard per-minute and per-message rates for the traffic itself. Direct Routing is the connection method, so Teams users keep the client they already have.
Can an AI agent framework drive calls directly?
Yes. An agent runtime can originate calls, attach to inbound calls, and read recordings and transcripts over the REST API, and receive call progress as webhooks. An agent that speaks the Model Context Protocol can treat those endpoints as tools. If you would rather not build the runtime, our managed AI Voice Agents bill at $0.10 per minute of talk time.
Something else? Contact us
Tell us what you run.
Bring the system you already have and we will tell you which of the three doors fits it.