Every Feature You Need
To Build AI Agents

Drag-and-drop node-based canvas for designing conversation flows and automation. Powered by React Flow with ELK.js auto-layout.

• 9 node types: Trigger, Action, Condition, Loop, Parallel, HumanInput, SubWorkflow
• Conditional branching (if/else logic)
• Loops and parallel execution
• Topological sort execution engine
• Execution trace panel for debugging
• Dry-run preview mode

Orchestrate teams of agents with 4 collaboration patterns. A supervisor delegates, agents execute in parallel or sequence.

• Supervisor-delegate orchestration
• Sequential task chains
• Parallel concurrent execution
• Dynamic routing by intent/context
• Team execution history and run tracking
• Agent performance metrics

Design guided customer journeys with conditional steps. Track user progress and detect off-path conversations.

• Visual pathway builder
• Node types: info, document, schedule, email, action
• Conditional branching
• User progress tracking
• Unmapped pathway detection

Comprehensive metrics on conversations, engagement, resolution rates, and user journeys. Funnel analysis and daily rollups.

• Conversation tracking with detailed metrics
• Funnel analysis with drop-off rates
• Active users and engagement charts
• Response time and latency histograms
• Conversation quality scoring
• Export and reporting capabilities

AI + rule-based sentiment tracking across conversations. Trend visualization over time to detect issues early.

• Per-conversation sentiment scoring
• Trend charts over time
• Alert on sentiment drops

Granular cost breakdown by model, agent, and feature. Set budgets, track spending trends, and get optimization recommendations.

• Per-model and per-agent cost breakdown
• Token counting and usage analytics
• Budget alerts and spending limits
• Cost trend visualization
• Model pricing reference

Full step-by-step traces of every agent turn. Waterfall visualization of LLM calls, tool executions, and timings.

• Waterfall trace visualization
• Per-step input/output inspection
• Timeline execution view
• Model and token breakdown per step

Bidirectional sync with HubSpot and Salesforce. Map fields, trigger syncs, and track sync history with conflict resolution.

• HubSpot and Salesforce connectors
• Custom field mapping UI
• Bidirectional sync with logging
• Connection testing and validation

Enrich leads with external data from Hunter.io, Apollo.io, and Clearbit. Single or batch enrichment with result tracking.

• Hunter.io, Apollo.io, Clearbit providers
• Single and batch enrichment
• Auto-enrichment on lead capture
• Results history and export

Capture, score, and manage leads from agent conversations. AI-powered lead quality classification.

• Lead capture from conversations
• LLM-based lead scoring
• Quality classification (hot, warm, cold)
• Lead detail and history views

Configurable safety rules to restrict topics, filter content, and enforce response limits. Built-in templates for common scenarios.

• Rule-based content validation
• Built-in guardrail templates
• Violation logging and alerts
• Pre- and post-response validation

GDPR data export, right-to-deletion, PII redaction, configurable retention policies, and immutable audit trails.

• GDPR data export and deletion
• LLM-powered PII redaction
• Data retention policies with cleanup
• Immutable audit log (who, what, when)
• Compliance configuration panel

Role-based access control with granular permissions per feature area. Custom roles, permission middleware, and role management UI.

• Owner, Admin, Member roles
• Feature:Action permission matrix
• Custom role creation
• Permission checking middleware
• Role management UI

Agents can respond with buttons, cards, carousels, embedded forms, data tables, and media. Template system for reuse.

• Quick reply buttons and chips
• Cards with image carousel
• Embedded forms for data collection
• Data table display
• Media player (image, video, audio)
• Response template system

Per-agent structured data tables with dynamic schemas. Agents can search, insert, update, and delete records during conversations.

• Dynamic schema definition
• CRUD via agent tool calls
• CSV import/export
• Filtering and query builder
• Persistent storage across conversations

Create, edit, and publish documents. Access control, public sharing via slug, and a template library.

• Document creation and rich editor
• Publishing and access control
• Public document URLs
• Template library

Publish, discover, and fork agent templates. Community ratings, reviews, search, and featured collections.

• Template publishing and discovery
• Community ratings and reviews
• One-click fork/install
• 9 default categories with search
• Featured and curated collections

Create dedicated workspaces for clients. Import agent configs, allocate credits, and manage settings per sub-account.

• Client sub-account creation
• Workspace switching
• Agent snapshot import/export
• Per-sub-account credit allocation
• White-label branding

Stripe-powered subscription management with a credit-based usage system. Promo codes, budget tracking, and customer portal.

• Stripe subscription management
• Credit-based usage metering
• Promo code support
• Customer billing portal
• Budget alerts and spending limits

Built-in AI assistants that help you build. Tool Builder, Pathway Builder, Value Model Assistant, and Console Helper.

• Tool Builder AI (generate tools from chat)
• Pathway Builder AI (design journeys)
• Value Model Assistant (pricing models)
• Console Helper (debug and navigate)

Agents learn and remember context from conversations. Per-user memory store with management UI for personalization.

• Memory extraction from conversations
• Per-user memory store
• Memory management UI (view, edit, delete)
• Bulk memory operations