Telli vs Callers (2026)

Agents run pre built or custom HTTP tools during the call, and telli integrates natively with Salesforce and HubSpot, calendars (Calendly, Cal.com, HubSpot Meetings, Zeeg, custom), Make, Zapier, n8n, Slack and Teams, with a contact lookup webhook for the customer's own systems.

Sourcetelli, docs.telli.com Tool Calls, Custom HTTP Tools and integration pagesread 2026-09-28

Scenario steps let the agent schedule a meeting, transfer the call or send an SMS during a call. After the call, actions email or text leads, update CRMs and launch workflows. Integrations cover Salesforce, HubSpot, Google Sheets, Slack, Cal.com and Zapier.

SourceCallers (formerly Voxia), docs.callers.ai Actions Summary and Campaign Scenariosread 2026-09-28

Visual workflows start on call ended, contact created or updated, a schedule or an incoming webhook and chain HTTP requests, contact updates, SMS, WhatsApp, email, CRM records and call scheduling, with If/else and Switch nodes for branching, beside an auto dialer with retry strategies.

Sourcetelli, docs.telli.com Workflows and Auto-Dialer pagesread 2026-09-28

An agent's script is a directed graph of script and nested script nodes joined by edges. It carries fallbacks for silence, errors, retries and unhandled input, along with tool functions, webhooks and scenarios that schedule meetings, transfer calls or send texts. Campaigns carry follow up cadences across call, chat and email.

SourceCallers (formerly Voxia), docs.callers.ai Create an agent API and Campaign Scenariosread 2026-09-28

Calls start from the auto dialer's schedules and retry strategies, API scheduled calls, inbound calls on telli or imported numbers, and workflows triggered by contact events, schedules and webhooks, with SMS and WhatsApp alongside voice.

Sourcetelli, docs.telli.com Auto-Dialer, Schedule Call, Phone Numbers and Workflows pagesread 2026-09-28

Triggers add contacts to a campaign automatically when something happens in a connected app, either instantly or on a five minute schedule, so calls start without anyone placing them. Inbound campaigns answer calls that customers start. Text, email and a web widget are further channels.

SourceCallers (formerly Voxia), docs.callers.ai Triggers, Starting a New Campaign and Create an agent APIread 2026-09-28

Knowledge bases of the customer's product guides, FAQs, policies and procedures (PDF, Word, Markdown, text) are searched by the agent during calls and re processed automatically whenever documents are added or removed, one knowledge base per agent of up to five documents (about 1,500 pages).

Sourcetelli, docs.telli.com Knowledge Base pagesread 2026-09-28

A knowledge base holds uploaded CSV, text, Word, PDF and Excel files, scraped website pages and typed text. Agents use it as context throughout calls, and customers keep it current by adding, editing or removing entries. Callers does not say how entries are retrieved during a call.

SourceCallers (formerly Voxia), docs.callers.ai Knowledge Baseread 2026-09-28

What carries between calls is the contact record and its custom properties, which agents read as variables and workflows update. That is the customer's contact data model, and there is no memory layer with its own scope and lifetime.

Sourcetelli, docs.telli.com Contacts, Contact Properties and Variables pagesread 2026-09-28

A Contact Memory store keeps memories per contact and campaign, each marked actual, outdated or deleted. API calls create, update or delete one memory, or clear all of a contact's memories. Callers does not say how memories are written or how long they last.

SourceCallers (formerly Voxia), docs.callers.ai Contact Memory APIread 2026-09-28

Each agent has a transfer number for complex issues or sales handovers and calls can be watched in live monitoring, but no person approves an agent action before it commits.

Sourcetelli, docs.telli.com Agents, Call Transfer and Live Monitoring pagesread 2026-09-28

Scenarios can transfer the call to a person, and the API plays a warm transfer summary to the receiving agent. Callers names no step where a person approves an agent action before it commits.

SourceCallers (formerly Voxia), docs.callers.ai Campaign Scenarios and API referenceread 2026-09-28

telli claims SOC 2, GDPR, HIPAA and EU AI Act compliance and EU hosting, and publishes a trust center at trust.telli.com. Settings cover inviting team members and API keys, but there are no roles, permissions or SSO.

Sourcetelli, telli.com homepage and docs.telli.com Settings pageread 2026-09-28

Single sign on supports SAML 2.0, OAuth 2.0 and OpenID Connect with Okta, Auth0 and Azure Active Directory. Callers says it is SOC 2 and GDPR compliant but names no report type. A trust center sits at trust.callers.ai. The older voxia.ai security policy attributes SOC 2 to its subprocessors, and no SOC 2 report of the company's own is published.

SourceCallers (formerly Voxia), docs.callers.ai Single Sign-Onread 2026-09-28

Each call record keeps the transcript, recording, an AI summary, a call score, goal attainment and the outcome, beside a performance dashboard and live monitoring, but the tool calls or actions the agent executed are not shown in the call record.

Sourcetelli, docs.telli.com Conversation History and Performance Dashboard pagesread 2026-09-28

Real time dashboards cover live and past campaigns with drill downs by agent, time and segment, and the API returns calls, recordings, statuses and each call's webhook payload. Callers names no per call view of the actions the agent took.

SourceCallers (formerly Voxia), docs.callers.ai Analytics and API referenceread 2026-09-28

Hosting is in the EU, by telli's account, a fixed region with no choice, dedicated or customer hosted option. The data retention settings govern how long call data is kept, not where.

Sourcetelli, telli.com homepage and docs.telli.com Data Retention pageread 2026-09-28

The service is hosted, and no hosting region is published. Callers names no region choice, private deployment or way to run it in the customer's own environment.

SourceCallers (formerly Voxia), docs.callers.ai privacy policy and documentation indexread 2026-09-28

Customers build agents in telli's no code builder with pre built in call actions and cookbooks, and use case pages describe lead qualification, reactivation, payment reminders and support, but no named packaged agents or agent templates are offered.

Sourcetelli, telli.com homepage and use case pages and docs.telli.com cookbooksread 2026-09-28

The customer builds each campaign by choosing inbound or outbound, naming the agent and describing the use case, and Callers then generates a script to edit. Callers names no prebuilt agents or template catalog.

SourceCallers (formerly Voxia), docs.callers.ai Starting a New Campaignread 2026-09-28

telli calls itself model agnostic across OpenAI, Google Gemini, ElevenLabs and Cartesia and lets customers pick a voice for each agent, but the thinking model cannot be changed, so telli chooses the LLM and only the voice is the customer's choice.

Sourcetelli, telli.com homepage and docs.telli.com Agents pageread 2026-09-28

An agent's configuration sets its voice, language and temperature, and customers can create or clone voices. No language model or voice provider is named, and no configuration field chooses among providers.

SourceCallers (formerly Voxia), docs.callers.ai Create an agent APIread 2026-09-28

A REST API covers agents, calls (get, list, initiate, schedule, batch), contacts and contact properties and phone numbers, with bearer token auth and HMAC signed webhooks for call ended, rescheduled, dialer and contact status events, and telli runs an MCP server at mcp.telli.com with OAuth.

Sourcetelli, docs.telli.com API reference, Authentication, Webhooks and MCP pagesread 2026-09-28

A public API at api.callers.ai has an OpenAPI specification and authenticates with an Authorization header. It creates and configures agents, including their script graph, channels, voice, tools, knowledge bases and webhooks. It also manages contacts, calls and recordings, webhooks, voices, contact memory and sub organizations.

SourceCallers (formerly Voxia), docs.callers.ai Create an agent API and API reference indexread 2026-09-28

Every call is scored for logical flow, clarity and satisfaction and marked on whether the agent reached its goal, scoring the agent's output after deployment, and refine your agent and feedback system guides explain how to act on those results. There is no test harness before release.

Sourcetelli, docs.telli.com Conversation History, Call Analysis and Feedback System pagesread 2026-09-28

Split Traffic runs concurrent versions of a campaign against live calls, and customers can report call issues for review. Callers does not say how the versions are compared or scored, and names no test cases or simulations.

SourceCallers (formerly Voxia), docs.callers.ai changelogread 2026-09-28

Agents act through HTTP tools, native integrations and workflows, and there is no browser, desktop or computer control. Navigating iOS call screening is a telephony behavior, not computer control.

Sourcetelli, docs.telli.com Tools and integration pagesread 2026-09-28

Agents work by phone, text, email and a web widget and act through integrations. Callers names no browser, desktop or computer control.

SourceCallers (formerly Voxia), docs.callers.ai documentation indexread 2026-09-28