AI Chatbot SaaS — Widget Demo

📋 Embed Code

<script
        src="https://www.365-chat.jp/static/widget/widget.js"
        data-api-key="sk_your_api_key"
        data-primary-color="#6C63FF"
        data-title="Chat with us"
        data-api-base="https://www.365-chat.jp"
        ></script>

🏢

Multi-Tenant

Each website gets its own namespace, API key, and isolated data. Register via POST /api/v1/websites/.

🔍

RAG Pipeline

Retrieval-Augmented Generation — answers are grounded in your ingested blog content via vector search.

📥

Blog Ingestion

Submit any URL or RSS feed. Content is scraped, chunked, embedded, and indexed automatically.

💬

Conversation Logging

Every message is persisted with token usage, context sources, and provenance for full auditability.

👍

Feedback Collection

Visitors rate responses 1–5. Data drives self-learning and fine-tuning pipelines.

⚡

Async Processing

Ingestion runs in background Celery workers. Track progress via the job status endpoint.

POST /api/v1/websites/ Register a new website tenant

GET /api/v1/websites/<id>/ Get website details & config

PATCH /api/v1/websites/<id>/ Update system prompt & LLM settings

POST /api/v1/chat/ Send a message through the RAG pipeline

POST /api/v1/ingest/ Submit URLs for background ingestion

GET /api/v1/ingest/<job_id>/ Check ingestion job status & progress

POST /api/v1/feedback/ Submit message feedback (rating + comment)