A VoIP provider for call centre traffic and diallers.

Traditional office telephony is often sized around users and concurrent conversations. A dialler starts attempts automatically, evaluates outcomes and can adjust its pacing continuously. This changes signalling behaviour and the need for connections available at short notice.

We therefore ask about campaign type, dialling mode, daily load, destinations, average call duration and expected peaks. These inputs are more meaningful than agent count alone. Personal lead content is not required for an initial technical review.

Preview dialling generally starts a call after an agent has selected the contact. Progressive dialling links the next attempt to an available seat. Predictive dialling forecasts availability and can create several attempts in parallel. The exact behaviour depends on your software and campaign configuration.

For the carrier, the important questions are how many new SIP dialogues begin per second and how many connections are ringing or connected at once. A seemingly small team can create a demanding load profile, while a large manually dialling team can be technically much quieter.

German fixed, mobile, European and international destinations are not interchangeable. Destination mix, time of day and volume influence routing and commercial terms. New countries or materially different volumes should be reviewed before the change goes live.

A second carrier can support comparison, migration or a defined fallback path. Routing in the dialler or SBC must remain unambiguous. Uncontrolled retries across multiple routes can distort results and complicate troubleshooting.

Caller IDs are checked against the right to use them and the intended callback path. From and P-Asserted-Identity are handled according to the agreed connection parameters. A visible number must not be mistaken for permission to spoof arbitrary identities.

Where campaigns use different numbers, format, selection and assignment are documented. This makes it possible to distinguish a technical fault from an unauthorised or incorrectly formatted caller ID.

ASR, ACD and PDD provide useful signals but do not replace case-level analysis. For a technical issue, timestamp, time zone, destination, SIP Call-ID and observed behaviour are especially useful. Audio and signalling are examined separately.

A pilot covers more than successful conversations. Busy, no answer, early media, DTMF, clean hangup and, where required, callback paths also belong in acceptance. Production load is enabled in a controlled manner after a bounded test.

SpeedNetwork presents technical handover points in Frankfurt at the Telehouse location associated with DE-CIX and in Berlin at the Level 3 location associated with BCIX. Depending on the project, connectivity can use the public internet, approved IP paths or separately agreed interconnection options.

The locations do not constitute a generic SLA or a statement that every customer automatically has a private circuit. Redundancy, preferred PoP and failover behaviour are agreed and tested in the service design.