Lemu
DocsLog inSign up

Lemu AI

Voice AI agents, designed as workflows

Build conversational agents in a visual editor—prompts, branches, and tools—then put them on phone lines or the web. Run outbound campaigns and see every conversation end to end.

Get started freeSee how it works

Platform

How it works

Connect context, design the conversation, and watch every run—from first prompt to closed call.

Orders table
CRM API
sync
Trigger ready
amount > 500
status = overdue
01

Connect your context

Pull in tables, APIs, or CRM fields so agents speak with the right customer data on every call.

Start

Start Call

→

Agent

Qualify

→

End

End Call

02

Design the conversation

Build Start, Agent, and End nodes with prompts and branches—the same visual canvas as production.

Run timeline

#1247
  • Trigger12ms
  • Condition8ms
  • Calllive
  • End…
03

Run and observe

Launch test or live calls, then inspect transcripts, step timing, and outcomes in one timeline.

Use cases

Built for real customer conversations

From collections to support queues, Lemu AI turns call scripts into visual agent workflows your team can ship and improve.

Collections agent

Live outbound · Invoice #4821

On call · 01:24

Live transcript

Calling about invoice 4821 which is past due…

Can I pay next Friday?

Sure — I'll schedule a reminder and send the link.

live

Collections & reminders

When an invoice crosses due date, trigger a workflow that looks up context, branches on amount, and sends a call or message.

NewRouteSLA

Support queues

Route tickets, notify Slack or email via HTTP nodes, and escalate when SLA timers expire—no manual triage.

POST /events→Agent

Partner webhooks

Ingest events from external systems, enrich with data lookups, and drive downstream actions in one graph.

Wait→Branch→Done

Ops playbooks

Wait, branch on conditions, schedule follow-ups, and end runs cleanly when the playbook completes.

Product

Everything in one voice platform

Agent flows, campaigns, tools, knowledge, and run visibility—one dashboard for designing and operating AI voice agents.

Agent flow

Active

Start

Greet

Agent

Qualify

End

Close

Runs today

128

Containment

86%

Automation workflows

DAG-based business automation with manual, schedule, webhook, event, and data_condition triggers.

Data layer

Tenant tables, Google Sheet sync, push APIs, and SQL—feed conditions and lookups in your graphs.

Automation runs

Per-run status, current node, context, errors, and step history for debugging and compliance.

Call step & voice agents

Optional AI voice in the graph via Call nodes, backed by conversational agent workflows when you need them.

Campaigns

Outbound calling campaigns with scheduling and contact lists when voice outreach is part of the playbook.

HTTP & tools

Call external APIs from automation nodes or attach reusable tools to voice agent flows.

Builder

Design agents visually

Map greetings, branches, tools, and endings on a canvas—same editor your team uses in production.

  • Visual agent canvas

    Start, Agent, Global, and End nodes with prompts and pathway labels—same editor your team uses in the dashboard.

  • Context-aware calls

    Attach tools, lookups, and customer data so agents respond with the right facts mid-conversation.

  • Test before production

    Run test executions and inspect step timelines so you validate paths before going live.

Start

Start Call

Continue

Agent

Feedback

End

End

End Call

Test run

Validate paths before going live

Start ✓
Feedback ✓
End · …

Launch your first voice agent

Sign up free, design a call flow in the visual editor, and test it before you go live.

Get started free

FAQ

Common questions about building with Lemu AI.

How do I get started?
Sign up, connect a data source or create a table, then build an automation workflow and activate it. Use test runs and Automation Runs to verify behavior.
What triggers are supported?
Manual runs, schedules, webhooks, events, and data_condition triggers that fire when rows match your rules.
How are runs tracked?
Every execution appears in Automation Runs with state, context, errors, and per-node history so you can debug without guesswork.
Can I still use voice?
Yes. Add Call nodes in automation workflows or build voice agent workflows for conversational steps—voice is optional, not required.
How is usage billed?
Organizations use a token wallet. Platform and voice consumption are tracked under Usage & pricing.
What auth options are supported?
Email and password sign-in, with optional Stack Auth for enterprise SSO and team management.