Delivering Our Strength
Voice platforms we build, host and run for you.
The Managed Outsource (TMO) owns a complete family of telecom software, from the carrier edge to the agent's headset. We deploy it on our own infrastructure, brand it as yours, and our team operates it day to day, so you get carrier-grade voice without building a telecom department.
- 7 products, one team
- Your brand on every portal
- Voices in 59 languages
The platform
Seven products that work together
Each product has its own web portal and runs on its own. Together they cover every hop a call takes: in from the carrier, through the switch, to an office phone, a softphone, a contact-centre agent or an automated voice campaign.
How a call flows through TMO
Languages
Speak to your customers in their language, anywhere in the world
Our IVRs, AI receptionist and voice campaigns speak to callers with text-to-speech. The voice engines behind them cover 59 languages, 47 of them with studio-quality HD neural voices. English and Swahili are live today, and any other language below can be switched on for your service.
Africa & Middle East
Europe
South Asia
East & Southeast Asia
Central Asia & Caucasus
Americas & Oceania (accents)
Voice engines: cloud HD neural voices (Google Chirp 3 HD), on-server Piper neural voices, and Whisper speech recognition. On-server voices keep audio inside your deployment. Cloud HD voices give the most natural sound.
Carrier & edge
Switching, routing and protection for operators
OneCoreOneCore · Wholesale softswitch
Wholesale switching with billing built in
OneCore is a class-4 softswitch for carriers, resellers and wholesale voice businesses. It sits between your customers and your upstream carriers. Every call is identified, priced and checked against the customer's balance before it connects, then sent down the cheapest working route.
How a call is handled
- Identify: the caller is matched by carrier IP, customer IP or SIP username and password.
- Rate & check credit: the destination rate is looked up, the balance is checked, and a maximum call duration is set from the credit available.
- Route: the call goes to carriers in least-cost order. If one rejects it, the next is tried automatically.
- Bill: the call is billed to the second, and it is cut off one check before the credit runs out, even when a customer has several calls running in parallel.
- Least-cost routingOrdered carrier routes per destination prefix with automatic failover.
- Real-time prepaid billingPer-second charging, credit checks and live cut-off.
- Carrier typesIP-authenticated carriers, and username/password carriers that OneCore registers to.
- Customer accountsIP-authenticated customers and registered SIP logins.
- DIDsRoute inbound numbers to customers or logins, with monthly DID fees charged automatically.
- Firewall & auto-banFailed-authentication bans with one-click unban.
- CDRsA-leg and B-leg records for every call.
- Live logs & SIP ladderSee signalling call by call for fast troubleshooting.
Technical data
- Signalling edge
- Kamailio
- Media
- rtpengine
- Call control
- FreeSWITCH (dynamic XML dialplan)
- Database
- PostgreSQL
- Protocols
- SIP over UDP / TCP
- Branding
- Product & company name set in the portal
OneSBCOneSBC · Session border controller
A border controller you can actually steer
OneSBC protects your voice network where it meets carriers, PBXs and the internet. Every route, trunk, number rule and codec policy lives in the database and is managed from the portal. The SBC looks it up on every call, so changes take effect immediately with no config files to edit.
What it does for each call
- Recognise the source by IP address, or by registration for endpoints.
- Pick the route: the first matching rule picks the trunk, and down trunks are skipped using live health pings.
- Rewrite: called and caller numbers, the From header and per-trunk SIP headers are adjusted to suit the carrier.
- Handle media: codecs are passed, preferred or transcoded, SRTP is offered, and DTMF is preserved.
- Fail over: on a carrier error such as a 503, the call tries the next trunk.
- Trunk healthOPTIONS pings mark trunks up or down automatically.
- Number & CLI rulesStrip, prepend and rewrite called and caller numbers.
- Header manipulationPer-trunk SIP headers, with From rewriting that is restored on replies.
- TranscodingG.729 ⇄ G.711 and more through media profiles. Verified live from a G.729 PBX to a G.711-only carrier.
- Topology hidingInternal addresses never leave your network.
- SRTP & TLSEncrypted signalling and media where the far end supports it.
- Carrier authenticationDigest authentication to carriers, plus outbound registration.
- Endpoints & DIDsRegistered devices with inbound numbers.
- ProtectionChannel limits and scanner detection with ban/unban.
- DiagnosticsCDRs with codecs per leg, SIP ladders and a live codec tester.
Technical data
- Engine
- Kamailio + rtpengine
- Transports
- SIP UDP / TCP / TLS
- Codecs
- G.711 µ/A-law, G.729, others via rtpengine
- Security
- SRTP, topology hiding, scanner bans
- Database
- PostgreSQL, looked up live per call
- Roadmap
- Full SIP manipulation engine, time-of-day routing, global CPS limits
MiniSBCMiniSBC · Compact border controller
A compact SBC for single sites and interconnects
MiniSBC is a lightweight FreeSWITCH-based border controller for a branch office, a single carrier interconnect, or a PBX, dialler or CRM that needs a safe path to the outside world. The portal is the single source of truth: it generates the gateways, access lists, dialplan, firewall rules and even the server's network configuration from its database.
Typical uses
- Connect an in-house PBX, dialler or CRM to one or more carriers.
- Terminate IMS / operator trunks that use P-Called-Party-ID.
- Share outbound traffic across carriers by weight, or fail over in order.
- TrunksCarriers and internal systems, with registration, P-Asserted-Identity, user=phone and tel: rewrite.
- Inbound routingBy DID, prefix range or IMS P-Called-Party-ID.
- Outbound routingBy caller ID and/or dialled-number prefix.
- Source matchingBy IP, or by SIP header regex for systems behind a shared NAT.
- Failover & load shareIn order or weighted, with strip/prepend per trunk.
- Smart retriesBusy, no-answer and unknown numbers are not retried on another carrier.
- FirewallDefault-deny nftables allowlist, kept in sync with the SIP ACL.
- NetworkInterfaces, routes and per-interface SNAT from the portal.
- MonitoringDashboard, live calls, CDR search, trunk and log watchdogs, and an audit log.
Technical data
- Engine
- FreeSWITCH 1.10 (mod_lua dispatch)
- OS
- Ubuntu 22.04 / 24.04
- Database
- PostgreSQL 14+
- Firewall
- nftables, default deny
- Network
- netplan-managed interfaces & SNAT
- Portal
- Python, HTTPS
Business voice
Phones for every office, desk and pocket
OnePBXOnePBX · Business phone system
The phone system your staff will actually like
OnePBX is a complete business PBX on Asterisk 22. Instead of static configuration, every call is routed live from the database. Ring groups, time conditions, IVR menus and call rules take effect the moment you save them. Administrators get one console; every staff member gets their own phone portal and browser phone.
What you can build
- Answer: an IVR or AI receptionist greets callers in a natural voice.
- Route: callers reach ring groups, queues or extensions, depending on the time of day and the caller.
- Handle: staff answer on desk phones, the browser phone or OneCall, with a caller pop-up showing a photo.
- Follow up: callers can leave voicemail, and the operator console handles transfers and parks.
- Visual IVR designerDrag-and-drop call flows with auto-layout, zoom and minimap.
- Text-to-speech promptsType a greeting and the system speaks it. Voices run locally (Piper), so no studio recording is needed.
- AI receptionistCallers say where they want to go and the system understands and routes them, using on-server speech recognition.
- Browser phoneA WebRTC softphone in any browser, plus a personal portal at /me.
- Operator consoleLive extension view with transfer, park, voicemail and hang-up.
- Call routingRing groups, queues, time conditions, inbound and outbound routes.
- Voicemail & pop-upsVoicemail, plus caller pop-ups with extension photos.
- Firewall & IDSZones, responsive SIP protection, intrusion detection and country blocking.
- Safe modeA firewall change that isn't confirmed rolls back in 90 seconds.
- ThemesNavy, Dark, White, Red or Green, per user or company-wide.
Technical data
- Engine
- Asterisk 22 + FastAGI live router
- Transports
- SIP UDP / TCP / TLS, WebRTC (WSS)
- Text-to-speech
- Piper neural voices, on-server
- Speech recognition
- Vosk, on-server
- Database
- PostgreSQL (incl. CDRs)
- Firewall
- nftables zones + IDS
OneCallOneCall · Softphone app
Your extension on Windows and Android
OneCall is TMO's own softphone, in the spirit of MicroSIP. The Windows desktop app and the Android app share one phone engine, so they behave the same everywhere. Staff can take their office extension home, on the road, or to a hot desk, without a desk phone.
How it connects
- OneCall registers over secure WebSocket (WSS) to OnePBX, OneBox or any WebRTC-ready PBX.
- It uses the same audio engine as Chrome, which brings Opus, echo cancellation and NAT traversal.
- For providers that only offer UDP 5060, a TMO SBC or PBX sits in between as the gateway.
- Up to 8 linesEach account registers on its own with a status light, like a desk phone.
- Line choicePick the outgoing line or set a default. Incoming calls show which line is ringing.
- Call controlMute, hold/resume, blind transfer, DTMF, and multiple calls with auto-hold.
- Per-account audioReorder codecs (Opus, G.722, G.711) and choose the DTMF method.
- Audio qualityEcho cancellation, noise suppression, automatic gain, and device choice.
- History & contactsDialpad, call history with a missed-call badge, and contact search.
- ProductivityAuto-answer, Do Not Disturb, dial prefix and five ringtones.
- ResilientRe-registers automatically with backoff after network loss.
- Windows extrasTray status, start with Windows, click-to-call sip:/tel: links, notifications.
- Android extrasSystem ringtone, earpiece/speaker toggle, proximity sensor and haptics.
Technical data
- SIP stack
- SIP.js (shared TypeScript core)
- Windows
- Electron + React
- Android / iOS
- Expo + react-native-webrtc
- Codecs
- Opus, G.722, G.711 µ-law / A-law
- DTMF
- RFC 2833, SIP INFO fallback
- Transport
- SIP over secure WebSocket (WSS)
Contact centre & automation
Talk to every customer, live or automated
OneBoxOneBox · Contact-centre suite
A complete contact centre in one box
OneBox runs inbound and outbound contact-centre operations: agents, in-groups, campaigns and self-service IVR. Its IVR can speak to callers, understand spoken answers, look the caller up in your own database, and hand the agent everything collected, so customers never repeat themselves.
A typical inbound journey
- The caller dials a DID and reaches a published IVR flow.
- The IVR checks business hours, then asks for an account number by keypad or speech.
- A data lookup queries your PostgreSQL, MySQL or SQL Server database and branches on the result.
- The call reaches the right in-group, and the agent sees the account, language and dialled number.
- Inbound & outboundIn-groups, DID routing and outbound campaigns.
- Drag-and-drop IVRPlay, menu, collect, hours, transfer, data lookup and condition steps.
- Text-to-speechPrompts are generated on the server, with variables like {account} spoken back to the caller.
- Speech recognitionCallers can answer menus by voice or keypad.
- English & SwahiliLocal voices for English (US and UK) and Swahili.
- SimulatorTest a flow before publishing it. Agents can dial *77 to try it live.
- Data sourcesPostgreSQL, MySQL and SQL Server connections for caller lookups.
- Agent contextIVR results are passed to the agent with the call.
- WebRTC agentsBrowser-based agent phones.
- SecurityFirewall with intrusion detection, and LDAP directory sign-in.
Technical data
- Engine
- Asterisk (AMI / ARI, PJSIP)
- Text-to-speech
- Piper: en-US, en-GB, Swahili
- Speech recognition
- faster-whisper, on-server
- Integrations
- PostgreSQL, MySQL, SQL Server
- Sign-in
- LDAP directory
- Portal
- Python (Flask), PostgreSQL, Redis
TTS AutodialerText-to-Speech Autodialer · Voice campaigns
Personalised voice messages, thousands of calls at a time
The Text-to-Speech Autodialer calls a list of customers and plays each one a message spoken in a natural neural voice, with their own details in it. Use it for payment reminders, collections, appointment reminders, alerts and announcements. No recording studio, no agents on the phone.
How a campaign runs
- Create: name the campaign, then set the institution, caller ID, language (English, Swahili or both) and voice gender.
- Write: type the message using variables: {name}, {institution}, {account} and {balance}.
- Upload: import the leads CSV. Rows with a missing name, phone, balance or institution, invalid numbers and duplicates are skipped and can be exported.
- Generate: the system builds each contact's audio from reusable segments. A message it has already generated is reused without another speech API call.
- Dial: calls go out in parallel within your schedule, and unanswered calls are retried.
- Neural voicesMale and female voices in British English and Kenyan Swahili.
- Bilingual campaignsEnglish, Swahili or both on the same call.
- PersonalisationName, institution, account and balance spoken per contact.
- Smart audio cachingRepeated phrases are generated once and reused, which cuts speech costs.
- Lead validationMissing fields, invalid phones and duplicates are caught at upload.
- Parallel diallingConcurrency limits keep trunks within capacity.
- Automatic retriesFailed or unanswered contacts are re-tried up to a limit.
- SchedulingGlobal or per-campaign calling hours and days, with automatic pause and resume.
- Live monitorWatch progress: answered, pending and failed (no answer, busy, early hang-up).
- ReportsDownloadable campaign reports, plus DIDs, trunks, users and test calls.
Technical data
- Dialler
- Asterisk call-file engine
- Voices
- Cloud neural TTS (en-GB, sw-KE)
- Audio
- Segment cache, ffmpeg conversion
- Scheduler
- Per-minute campaign scheduler
- Status
- Real-time call status updater
- Portal
- PHP + MariaDB
Why a managed outsource
Our strength, delivered to your business
Every product on this page is our own software. When you need a feature, we can build it, instead of waiting on a vendor roadmap.
Hosting, firewalls, intrusion detection, monitoring, backups and upgrades are handled by our team. You focus on your customers.
Your name, logo and company details appear on every portal. They're settings, not code changes.
Get in touch
Let's talk about your voice stack
Tell us what you run today, whether that's carriers, offices, a contact centre or collections, and we'll show you the platform live on our servers.
Request a demo