2 modes
realtime or pipeline handling depending on the job
AI Agents
Use AI agents as part of operations, not as a side experiment.
Momo Pulse lets teams configure AI agents with voice, model, instructions, and tools, then tie them back into the same call and routing environment used by the rest of the team.
Realtime modePipeline modeTool syncCall-aware deployment
Support Agent
Realtime mode · Active
Live conversation1:24
Caller
I need to check on my refund for order 2847.
Using tool: check_order_status
AI Agent
Your refund of $42.00 for order #2847 was processed yesterday. It should arrive in 2-3 business days.
Active tools
HTTP API Tickets MCP N8N Workflow
Voice: Coral
Synced
Tool ready
HTTP and prompt-style tool payload support
Call linked
AI stays attached to real call operations
How AI agents work
From customer contact to AI-resolved outcome
Inbound
Call, WhatsApp, or IVR dispatch
CallWhatsAppIVR
AI Agent
Listens, reasons, and acts with tools
STT → Understand intent
Execute tools (HTTP, MCP)
TTS → Voice response
Tool Layer
HTTP, MCP servers & N8N workflows
HTTP API calls
Tickets MCP server
N8N automations
AI resolved
Voice reply + tool action
Closed1m 24s
Ticket created
Via MCP server
HIGHAssigned
Human handoff
Forwarded to ring group
Transferred
Realtime Voice Pipeline STT/TTS Tool Orchestration MCP Servers Call Routing
What teams unlock
Configurable AI that fits real customer flows.
AI agents inherit the surrounding context of the platform, which keeps them grounded in calls, tools, routing, and operator review.
Realtime or pipeline delivery
Choose the AI mode that matches the workflow and the latency or orchestration profile you need.
- Realtime voice option
- Pipeline speech workflow
- Configurable voice and model
Tool orchestration
Attach tools so the AI can act on customer intent instead of only generating answers.
- HTTP call tools
- Prompt-driven tool support
- Schema-based parameters
Operational sync
Keep AI agent configuration connected to the tenant and to external synced configuration where available.
- External agent config IDs
- Per-agent call views
- Configuration and tool tabs
1
Define behavior clearly
Set mode, voice, model, instructions, and tools so the agent has an operational role instead of vague intelligence.
2
Assign the agent into the workflow
Apply the agent to numbers and routes where AI coverage should take first pass or overflow traffic.
3
Review and adjust from live outcomes
Inspect call history, tool configuration, and synced state from the same product surface.
Why teams choose it
AI is treated as a managed operator layer
That means it has explicit configuration, tool definitions, and call history instead of being isolated from the rest of the product.
- Agent mode, voice, and model controls
- Tool payload normalization and sync
- Call statistics attached to each AI agent
Operator lens
Balanced design, low visual noise
The interface style follows the logo palette without turning every surface into a neon gradient. Most UI stays bright, neutral, and operational, while brand color is used for focus and motion.
Primary tone
Blue first, pink reserved for emphasis
Motion
Slow float and reveal, not constant noise
Images
Custom in-repo SVG illustrations
CTAs
Clear path to demo or workspace creation
Ready to add AI coverage?
Deploy AI agents that stay connected to your real routing and operator flow.
Use AI where it helps, keep the controls visible, and expand human capacity without losing oversight.
AIToolsCall routing