Streaming chat
Deploy conversational agents with AI SDK streaming. Every reply grounded in your uploaded context.
Multi-agent voice + chat platform
Build agents that speak and listen
Create, test, and deploy AI agents with chat, voice, and document context. Ground every response in your knowledge base — powered by Gemini and Neon pgvector.
Real-time voiceRAG-powered answersOne-line embed
Agents
Support Bot
Sales Assistant
Onboarding
Context
pricing.pdf
faq.md
Hi! I'm your support agent. I can answer questions about pricing, features, and onboarding — grounded in your docs.
What's included in the Pro plan?
Pro unlocks voice mode, unlimited context files, additional agent seats, and production deploy via API keys and embed widget.
Context
guide.docx
onboarding.mp4
policies.pdf
Retrieved from pricing.pdf
Ask anything about your product…
Features
Everything you need to ship AI agents
From sandbox to production in one platform. Chat, voice, retrieval, and deployment — without stitching together five different tools.
Live voice
Gemini Live over WebSocket. Talk naturally in the playground, then ship voice to production.
Document RAG
PDF, DOCX, markdown, images, audio, and video — chunked, embedded, and retrieved via pgvector.
Multi-agent
Run separate agents for support, sales, and onboarding — each with its own prompt and knowledge.
Developer-first deploy
REST chat API, WebSocket voice, and a one-line embed widget. API keys with losono_sk prefix.
Custom system prompts
Fine-tune personality and boundaries per agent
Sandbox mode
Test safely before publishing to production
Conversation logs
Review every interaction and track usage
How it works
From idea to live agent in minutes
No infrastructure to manage. No vector database to configure. Just build, test, and ship.
01
Create your agent
Sign in with Google, name your agent, and write a system prompt that reflects your brand voice.
02
Upload context
Drop in docs, PDFs, and media. Losono chunks, embeds, and indexes everything in Neon pgvector.
03
Test in playground
Chat and talk to your agent in sandbox mode. Preview retrieved context before going live.
04
Publish & deploy
Flip the switch, generate API keys, and embed a widget on your site — or integrate via REST.
Pricing
Start free, scale when you're ready
No credit card required to begin. Upgrade to Pro when you need voice, unlimited context, and production deployment.
Free trial
$0 forever to start
Perfect for exploring the platform and building your first agent.
- 1 chat-only agent
- Up to 3 context files (10 MB each)
- Playground testing
- Document RAG with pgvector
- Custom system prompts
Pro
Per seat billed monthly
Everything you need to ship voice agents and embed them in production.
- Voice playground & deployed voice
- Unlimited context files per agent
- Additional agent seats
- Embed widget & API keys
- Conversation logs & usage tracking
- Production publish mode
Developers
Ship with three lines of code
REST chat with streaming, WebSocket voice, or a drop-in embed widget. Full API reference in the docs.
Read the API docsEmbed widget
<script src="https://losono.xorora.com/embed.js" data-agent="your-agent-slug"> </script>
Chat API
POST /api/agents/{id}/chat
Authorization: Bearer losono_sk_...
{ "messages": [...] }Ready to build agents that actually know your product?
Join developers shipping voice and chat agents with grounded, retrieval-augmented responses — in minutes, not months.