Talenoa
The Run

What actually happensbetween the listand the outcome.

Every platform shows you the call. Almost none show you the machinery around it — the suppression rules, the retries, the finalizer, the writeback receipt. This is the whole run, stage by stage, exactly as we operate it for you.

Station 01/Import

The right batch, not the whole book.

The run starts before any dialing. Our operators pull leads from your CRM with the filters that matter — cohort, expiry window, territory, calling hours — plus the retry policy and campaign defaults. The batch is previewed before a single number is queued.

On DIY platforms this is a CSV upload and a prayer — no filters, no preview, no calling-window rules.

Source
CRM query, not a spreadsheet
Scoped by
cohort · window · territory · hours
Before dialing
full batch preview

Station 02/Suppress

The calls that never happen.

The most important calls in the run are the ones we refuse to place. Duplicates, numbers already inside an active campaign, accounts under a no-repeat window, and leads a human already owns are pulled out in preview — and checked again in the seconds before each dial.

Skip this and the same customer gets three robots in one week. It is how AI calling gets a bad name.

Rejected
duplicates · active-campaign overlap
Protected
no-repeat windows · human-owned leads
Checked
at preview, then again pre-dial

Station 03/Strategy

A plan per lead, decided before the ring.

Each lead's signals — tier, history, language, lapse reason — pick the opening hook, the pitch depth, and the objection routes before the phone rings. A lapsed high-value account and a brand-new enquiry get entirely different conversations from the same campaign.

One prompt for every caller means the new enquiry gets pitched like a lapsed account. Both hang up.

Read
tier · history · language · lapse reason
Decided
opening hook · pitch · objection routes
Result
no two leads hear the same call

Station 04/The call

Live, grounded, and allowed to act.

The agent speaks the caller's language and switches when they do. Mid-conversation it can search your documents, hit your live APIs, resolve catalog items, send an approved WhatsApp follow-up, or hand the call warm to a human — every action logged as it happens.

A voice demo is easy. A voice that can check a warranty API mid-sentence and log the evidence is the hard part.

Language
detected and switched mid-call
Can act
knowledge · APIs · WhatsApp · transfer
Every action
logged with its result

Station 05/Finalize

One official outcome per call.

When the call ends, the finalizer commits a single authoritative disposition — backed by the transcript and the tool evidence, with a confidence score. Not a summary paragraph. A structured outcome your systems and your team can act on.

Most platforms hand you a transcript and call it an outcome. Someone still has to read it.

Commits
one disposition, evidence-backed
Carries
reason · confidence · objection tags
Format
structured, not prose

Station 06/Work

The call creates the next step, not a note.

A promise date becomes a scheduled reminder. A hot lead becomes an owned task with a due time. An unresolved issue becomes a priority callback. The run's output is work sitting in a queue with a name on it — never a note someone might read.

Without this stage, every outcome dies in a notes field. Follow-up is where DIY setups quietly fail.

Creates
tasks · callbacks · reminders
Each carries
owner · due time · call context
Never
a note in a field nobody reads

Station 07/Writeback

Your CRM finds out without being asked.

The disposition, the enriched fields, and the lifecycle move are queued back to your CRM or webhook the moment the outcome is final — without blocking the call record. Vendor hiccups retry automatically; nothing is lost between systems.

This is the integration project that stalls DIY rollouts for months. Here it ships configured.

Synced
disposition · fields · lifecycle stage
Via
CRM writeback or webhook
On failure
queued and retried, never dropped

Station 08/Audit

Every outcome can defend itself.

Transcript, tool calls, finalizer reasoning, writeback receipt — the full chain is kept for every call. When someone asks “why is this account marked renewal-interested?”, the answer is on file, timestamped, down to the sentence.

The question isn't whether someone will ask. It's whether you'll have the answer.

On file
transcript · tools · reasoning · receipt
Answers
why this outcome, from which evidence
Kept
timestamped, per call, always
The whole run

You saw eight stations. Your team sees one: the outcome.

Everything above is our job — configured, monitored, and improved by the Talenoa team. Your job is the first conversation, where you tell us what a good outcome looks like. The run is live on your leads within 48 hours.

You operate
nothing
You watch
outcomes landing in your CRM
We run
all eight stations, around the clock