Siyazen
Get started
Intelligent Communication

Turn your documents into an AI chatbot that only answers from what you gave it

Upload your policies, FAQs and manuals. Siyazen extracts, indexes and grounds every reply in them — embedded on your site, scripted as a pre-chat flow, or called from your own backend. Priced in rupees, hosted in India.

Get started See what it does
₹1 per answerData in AWS ap-south-1Built for DPDPNo AI team required
Ask about our refund policy
Do refunds cover international orders?
Yes — international orders are refundable within 14 days of delivery, per our Shipping & Returns policy. Duties already paid aren't refunded.

Answered from the merchant's own policy PDF

Type your question… ↵
What it does

Ten capabilities, one shared pipeline

A dashboard to manage documents, an embeddable widget for your visitors, and a server-to-server API for developers — all fed from the same document-grounded index.

Core

Document-grounded chat

Hybrid dense + full-text retrieval over your own library, with a semantic answer cache so repeat questions cost nothing.

Ingestion

Any document, not just clean PDFs

PDF, DOCX, TXT and images. Scanned pages fall back to OCR; diagrams and photos are described by a vision model so they're searchable too.

Distribution

Embed anywhere

A draggable chat bubble via one script tag, or a plain iframe for CSP-locked sites. Branded to match your site.

Conversion

Visual flow builder

A drag-and-drop canvas for the pre-chat sequence — greeting, qualifying questions, lead capture — before handing off to AI chat.

Safety

Guardrails, authored or AI-suggested

Rules that shape what the bot will and won't say, with an assistant that proposes rules from your own documents, quoted evidence included.

Transparency

Cost you can see

A prepaid wallet with a live balance and a flat, published per-message price — no surprise invoice.

See all ten capabilities →
Mechanism

How it works, in two flows

Getting knowledge in

  1. Upload file
  2. Store in S3
  3. Extract text (native / OCR / vision)
  4. Chunk
  5. Embed
  6. Index in pgvector

Answering a question

  1. Visitor asks
  2. Moderation check
  3. Semantic cache lookup
  4. Retrieve top chunks
  5. Apply guardrails + history
  6. Generate answer
  7. Log + bill
Pricing

Two meters. No surprise invoice.

Monthly subscription
₹500 / month

Covers ingestion — OCR, storage, embedding every uploaded document.

  • Up to 25 documents indexed per 30-day period
  • Required before any upload or chat reply
Pay-per-answer wallet
₹1 / answered message

Prepaid balance; a flat, published price instead of a metered bill.

  • Semantic-cache hits and greetings are free
  • No expiry on wallet balance
See the full pricing breakdown →

Activate your assistant

Connect your documents and have a working chatbot in minutes.

Get started Request a demo