New carrier onboarding checklist for wholesale VoIP

Bringing a new carrier onto your network — whether as a customer sending traffic to you or as a vendor you send traffic to — is one of the most common jobs in wholesale voice, and one of the easiest to get slightly wrong. A missed IP address means failed calls on day one; a missing credit limit means unlimited exposure; a forgotten test means the first person to discover a broken route is a paying customer. This checklist walks through a repeatable onboarding process, from the commercial paperwork to the first week of live traffic.

Why a checklist matters

Interconnects are usually set up under time pressure: sales has signed the deal and the partner wants to send traffic today. Under pressure, people skip steps they think are optional. A written checklist turns onboarding into a routine that any engineer on the team can follow and any manager can audit. It also creates a record of what was agreed, which is invaluable later when there is a dispute about rates, billing increments or who changed what.

The checklist below is split into phases. Each phase has a clear owner and a clear “done” state. Adapt it to your own business, but resist the temptation to remove the steps that feel boring — they are the ones that prevent expensive mistakes.

Phase 1: commercial and compliance

• Signed agreement covering services, payment terms, billing increments, dispute windows and fraud liability.
• Company verification: registered company details, a real business address, and named commercial, technical, billing and NOC contacts.
• Payment model: prepaid, postpaid with a credit limit, or a mutual netting arrangement. New partners should normally start prepaid or with a small limit.
• Traffic profile: expected destinations, volumes, peak concurrency, calls per second, and traffic type (retail, CLI, CC). A clear profile makes monitoring meaningful later; CLI vs NCLI routes explained covers why traffic type matters.
• Compliance: confirm the partner’s authorisation to send or carry traffic where required, and agree caller ID rules for each destination.

Done when: the agreement is signed and every field on your partner record is filled in.

Phase 2: technical exchange

Exchange a short, structured technical form rather than a long email thread. A typical form includes:

Send your own details in the same format: your signalling and media addresses (for example 198.51.100.20), your supported codecs and your number format. The codecs and capacity lesson helps you check that the agreed channels fit your bandwidth, and Lesson 13 on number formats explains how to normalise whatever format the partner sends.

Done when: both sides have confirmed each other’s technical details in writing.

Phase 3: security

• Add the partner’s signalling (and, if needed, media) addresses to your firewall allow-list — and nothing wider.
• Configure the softswitch to accept traffic only from those addresses on the partner’s gateway.
• Avoid username/password registration for wholesale interconnects; if it cannot be avoided, use strong unique credentials combined with an IP restriction.
• Apply default destination blocks for high-risk ranges, as described in our guide to IRSF and toll-fraud protection.

The underlying principles are in Lesson 10: VoIP server security.

Done when: the partner can reach only what they need, and nothing else.

Phase 4: switch configuration

On VOS3000 and similar softswitches, a new partner usually involves several linked objects. Build them in a consistent order:

1. Rate group with the agreed rates, billing increments and effective date. Double-check the prefixes against the partner’s deck.
2. Account with the payment type, balance or credit limit, and daily spend limit.
3. Mapping gateway (for a customer) or routing gateway (for a vendor) with IPs, codecs, channel and CPS limits, and any prefix manipulation.
4. Routing: add a new vendor to the routing for the agreed prefixes at low priority first, so it does not take production traffic before testing.
5. Naming: use a consistent naming convention so the NOC can recognise the partner in reports at a glance.

A full walk-through of each object is in Lesson 4: VOS3000 accounts, gateways, prefixes and rate groups, with background in Lesson 3: VOS3000 overview and architecture.

Done when: a second engineer has reviewed the configuration against the technical form.

Phase 5: testing before go-live

Never let real traffic be the first test. Agree a test window with the partner and work through:

• Signalling: an OPTIONS ping or test INVITE reaches the other side and gets an expected response.
• Test calls to handsets you control in each main destination. In test plans and written examples, use reserved test numbers (for example the +44 20 7946 0xxx drama range) rather than real subscriber numbers.
• Two-way audio on every test call, checking both directions.
• CLI delivery: compare the number shown with the number sent.
• DTMF and fax if the traffic needs them.
• Billing: confirm the test calls appear in CDRs with the correct rate, duration and increment.
• Failure handling: call an invalid number and check the response code is sensible, so failover behaves correctly.

If anything fails, Lesson 9: troubleshooting common VoIP issues walks through the usual suspects. Lesson 8: route testing methods covers how to build a repeatable test plan.

Done when: every test is recorded as passed, with the time and the engineer who ran it.

Phase 6: go-live

• Tell the partner and your NOC the exact go-live time.
• For a vendor, raise its priority gradually — start with a share of traffic on a few prefixes, then widen.
• For a customer, start with the agreed initial limits, not the long-term target.
• Watch the first hour live: failure codes, ASR, ACD, PDD and audio quality.

Phase 7: the first week

The first week is when most onboarding problems show up. Plan for it:

• Compare ASR and ACD with the expectations agreed in phase 1; how to check VoIP route quality explains which metrics to read together.
• Watch for False Answer Supervision on new vendors: unusually high ASR with short ACD is a warning sign.
• Reconcile a sample of CDRs against the partner’s records after the first day; see Lesson 5: rates, billing, balance and CDRs.
• If metrics fall, follow a structured approach such as our guide to debugging ASR and ACD drops on VOS3000.
• Review limits at the end of the week and raise them only if traffic and payments justify it.

Common onboarding mistakes

• Whitelisting a whole range “to be safe” when the partner has two signalling addresses.
• Leaving a new vendor at top priority before any test call.
• Forgetting the media IP range, which causes one-way or no audio.
• Setting a billing increment in the switch that differs from the contract.
• No credit limit on a postpaid customer because the deal was “urgent”.
• No written record of who configured what and when.

One-page onboarding checklist

1. Agreement signed; contacts, payment model and traffic profile recorded.
2. Technical form exchanged and confirmed by both sides.
3. Firewall and switch allow-lists set to specific addresses.
4. Rate group, account, gateway and routing configured and peer-reviewed.
5. Limits set: channels, CPS, credit and daily spend.
6. Test calls passed for signalling, audio, CLI, billing and failure handling.
7. Gradual go-live with the NOC watching.
8. First-week review of KPIs, CDR reconciliation and limits.

New to some of the terms? The wholesale VoIP glossary and Lesson 1: wholesale carrier basics are good starting points.

Interconnect with VoxComms

VoxComms follows this process for every new interconnect, with NOC testing before and after go-live. For carriers on VOS3000, VoxInspector makes the first-week review easy with per-partner ASR, ACD, PDD and MOS, FAS detection and route testing. To start an interconnect, contact our sales team.