Skip to content

AI Agent Builder

Build and deploy specialized AI agents without writing code.


What It Is

The Opensure Agent Builder is an AI agent development platform built into the Phoenix dashboard. Deploy specialized agents—RAG assistants, lead qualifiers, voice bots, content generators—in minutes, not months.

No coding required. Bring your own API keys. Own your data.


Who It's For

  • Solo Brokers: Automate client interactions and qualify leads 24/7
  • Agency Teams: Scale operations without scaling headcount
  • Operations Managers: Centralize knowledge access and automate compliance workflows

Agent Types

RAG Agent (Knowledge Base)

Upload your documents. Ask questions in plain English. Get answers with source citations.

What it handles:

  • Policy lookup and coverage queries
  • Procedure and compliance FAQs
  • Onboarding documentation
  • Client-facing knowledge bases

How it works:

  1. Upload PDFs, DOCX, or TXT files
  2. Documents are chunked and embedded automatically
  3. Ask questions via chat interface
  4. Responses cite specific sources

Lead Generation Agent

Qualify incoming leads automatically with customizable scoring criteria.

What it handles:

  • Initial lead qualification
  • Scoring based on your criteria
  • Auto-responses to common questions
  • CRM integration

Voice Agent

Answer calls, route inquiries, and handle common questions via phone or web.

What it handles:

  • Inbound call handling
  • Appointment scheduling
  • Claims intake
  • After-hours support

Integration with Vapi or Retell AI.

Content Creation Agent

Generate personalized emails, proposals, and marketing content.

What it handles:

  • Client follow-up emails
  • Renewal reminders
  • Quote proposals
  • Marketing campaigns

Key Features

No-Code Deployment

Deploy agents directly from the dashboard. Configure templates, upload knowledge, and go live—all without touching code.

BYOK (Bring Your Own Key)

Connect your own API keys for Claude, OpenAI, or other LLM providers. No vendor lock-in. Full cost control.

Supported providers:

  • Anthropic Claude
  • OpenAI GPT
  • Additional providers coming soon

Insurance-Native Templates

Purpose-built templates for insurance workflows. Agents understand your terminology out of the box:

  • Policy lookup patterns
  • Claims intake flows
  • Carrier matching
  • Quote generation

Unified Data Layer

Agents connect directly to your Opensure data—clients, policies, carriers. No separate integrations. No data sync headaches.

Multi-Tenant Isolation

Your data stays yours. Agents only access data within your tenant scope. No cross-tenant data leakage, ever.


Data Sovereignty

Customer data never enters Opensure infrastructure. Agent processing happens within your data scope:

  • Documents stay in your knowledge base
  • LLM calls use your API keys
  • Conversation logs are broker-scoped
  • Export and delete your data anytime

Getting Started

1. Add Your API Keys

Navigate to Settings > API Keys in the dashboard. Add your Claude or OpenAI key. Keys are encrypted at rest.

2. Create an Agent

Go to Agents > New Agent. Pick a template or start from scratch:

  • RAG Agent: For knowledge base Q&A
  • Lead Agent: For lead qualification
  • Voice Agent: For phone/web calls
  • Content Agent: For email/proposal generation

3. Configure

Customize prompts, upload documents, set scoring criteria—whatever your agent type requires.

4. Deploy

Hit deploy. Your agent is live. Monitor performance from the dashboard.


Template Library

Pre-built templates for common insurance use cases:

TemplateUse Case
Policy LookupFind coverage details from uploaded policy docs
Claims FAQAnswer common claims questions
Renewal AssistantGenerate renewal reminder sequences
Quote BuilderCompile quote proposals from templates
Carrier MatcherMatch risks to likely-to-quote carriers
Compliance CheckerValidate procedures against regulations

More templates added regularly based on user feedback.


Analytics & Monitoring

Track agent performance from the dashboard:

  • Query volume: How often agents are used
  • Response time: Average latency per interaction
  • Cost tracking: Token usage and API spend
  • Success metrics: Resolution rates and user satisfaction

Architecture

┌─────────────────────────────────────────────────────────────┐
│                      Your Dashboard                        │
│  ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐       │
│  │RAG Agent │ │Lead Agent│ │Voice     │ │Content   │       │
│  │          │ │          │ │Agent     │ │Agent     │       │
│  └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘       │
│       │            │            │            │              │
│       └────────────┴─────┬──────┴────────────┘              │
│                          │                                  │
│                   ┌──────▼──────┐                           │
│                   │ Agent Engine │                          │
│                   │  (LangChain) │                          │
│                   └──────┬──────┘                           │
│                          │                                  │
│           ┌──────────────┼──────────────┐                   │
│           │              │              │                   │
│    ┌──────▼────┐  ┌──────▼────┐  ┌──────▼────┐              │
│    │Your LLM   │  │Knowledge  │  │Your Data  │              │
│    │API Keys   │  │Base       │  │(Clients,  │              │
│    │(BYOK)     │  │(Vector DB)│  │Policies)  │              │
│    └───────────┘  └───────────┘  └───────────┘              │
└─────────────────────────────────────────────────────────────┘

Security

  • Encrypted storage: API keys encrypted with Fernet at rest
  • BYOK architecture: You control LLM provider access
  • Broker-scoped queries: All data access filtered by tenant
  • Audit logging: Every agent interaction logged
  • HIPAA-ready: Voice agents support compliance requirements

Current Status

The Agent Builder is in active development as part of the Q1 2026 roadmap.

Phase 1 (MVP): RAG Agent with knowledge base upload, BYOK support, and chat interface.

Coming next: Lead generation agents, voice integration, content creation, and multi-agent workflows.


What's Next

  1. Explore the Dashboard: app.opensure.dev
  2. Add API Keys: Configure your LLM provider credentials
  3. Create Your First Agent: Start with a RAG knowledge base

Questions? Check the Help Center or contact support.

Built with VitePress