Open infrastructure

Connect your data without a complex project

Informio respects your existing systems. Supply data in a clear format and publish with one script.

Informio Search

Your assistant knows your content. Help people find it.

Add autocomplete, AI search results and links to your pages or products. Embed it with a single script.

JavaScript widget

One embed tag, Shadow DOM isolation, allowed domains, and responsive behavior.

Informio XML 1.0

Snapshot or incremental feed, scheduling, ETag, and change hashing without AI tokens.

PrestaShop module

Install the module directly in your store. It adds the widget, an Informio XML feed for CMS pages, categories, FAQs and blog posts, plus a protected order status feed.

Download module

Documents

Secure PDF, DOCX, and TXT upload with asynchronous extraction and indexing.

OpenAI or custom AI server

Encrypted access key, model selection, endpoint security test, and audit for every operation.

Dokly invoicing

Proforma invoices, invoices, payments, webhooks, and automatic plan renewals.

API-ready architecture

AI provider, vector store, and ingestion are separated for future integrations.

Slack alerts

Deliver new chats and live operator requests to your team with context and a link to the correct conversation.

AI on your terms

Start without an API key. Change the infrastructure whenever you need.

Every assistant uses secure Informio AI by default. If you need your own billing, models, or full control over the data path, connect your OpenAI account or OpenAI-compatible infrastructure.

With default Informio AI, the tokens required for answers and knowledge processing are included in the monthly fee within your plan's answer limit. There is no separate token charge.
Default

Informio AI included

Works automatically without an OpenAI account, API key, or separate credit. Ideal for a fast launch.

Your own OpenAI account

A BYOK connection for companies that want their own project, models, and direct OpenAI billing.

Your own AI server

Connect a cloud or self-hosted OpenAI-compatible server with endpoints for models, responses, and embeddings.

Once a working custom connection is saved, Informio automatically gives it priority. Without one, assistants safely continue through Informio AI. Costs of your own provider or server are billed separately.
For e-commerce · Heureka XML

Turn your existing product feed into a live knowledge base

Connect the same Heureka-compatible XML feed your store already uses. Informio imports products, prices, categories, and parameters, while an availability feed adds exact stock and delivery times.

No need to rebuild your existing export

We process a standard SHOP/SHOPITEM feed directly and match products using stable ITEM_ID values.

Price and stock updates without wasted AI tokens

Commercial data updates separately. New embeddings are created only when a name, description, or parameter changes.

Automatic feed quality score

See how many products lack descriptions, parameters, EANs, images, prices, or availability—and why that limits assistant answers.

Product source

E-commerce catalog

Heureka XML
Data quality for the AI assistant
87 %Excellent quality
1,248 products
Product descriptions76 %

Add descriptions to 24 products.

Technical parameters98 %

Parameters are ready for accurate comparisons.

Scheduled differential sync without reprocessing the entire catalog

Google product catalogue

Google Merchant RSS 2.0
<g:id>ABC-1048
<g:title>Wooden rosary
<g:price>39.90 EUR
<g:availability>in_stock
<g:product_detail>Material: wood

1,248

products synchronized

Live

prices and availability

0

embeddings for a price change

For e-commerce · Google Merchant feed

Turn your Google product catalogue into current assistant knowledge

Connect the same RSS/XML feed you already send to Google Merchant Center. Informio regularly imports products, regular and sale prices, availability, categories, images, and custom attributes.

No additional export required

We support standard RSS 2.0 feeds with Google g: attributes and a stable product g:id.

Prices and availability stay live

Commerce changes are stored without new embeddings. AI content is reindexed only when a name, description, category, or attribute changes.

See immediately what the feed is missing

The quality score highlights missing descriptions, attributes, EANs, brands, images, prices, or availability.

Live knowledge through XML

Your AI works with data that updates together with your system

Publish products, services, prices, availability, or notices in Informio XML 1.0. Informio checks the feed on schedule and automatically synchronizes the assistant's knowledge.

01

Your system publishes XML

Your store, ERP, or CMS provides items with stable IDs and agreed fields.

02

Informio checks the feed

Synchronization runs on your chosen schedule, from every 15 minutes to daily.

03

Only differences are compared

Content hashes detect new, changed, and removed items without calling AI.

04

Knowledge updates automatically

Only new and changed parts are indexed. Unchanged knowledge remains untouched.

Always-current information

Price or availability changes reach the assistant's knowledge after synchronization.

Answers grounded in your data

The assistant searches approved knowledge and admits when information is unavailable.

No wasteful reprocessing

The entire catalog is not embedded repeatedly. AI processes only content that actually changed.

Efficient synchronization

Where tokens are used — and where they are not

Checking XML is a technical operation. AI is involved only when embeddings are created for new or changed content and when answering a visitor.

Download and validate XML0 AI tokens
SHA-256 item comparison0 AI tokens
Knowledge indexingNew and changed parts only
Answering a visitorNormal usage of the selected model

What does it mean that the assistant does not make things up?

Before answering, Informio retrieves relevant parts of your knowledge base and instructs the model to answer only from them. If the evidence is insufficient, the assistant should say that the information is unavailable.

No generative model can honestly be described as absolutely error-free. Informio substantially limits guessing through controlled sources, RAG retrieval, testing, and answer auditing.

Live handoff · Slack

When a customer needs a person, your team learns about it where they already work

Informio sends a new-chat alert or a prominent operator request to the selected Slack channel. Your team does not need to watch the administration all day.

Configure two events separately

A subtle alert for the first message and a priority card when the visitor explicitly requests a person.

Operators do not open chats blindly

Slack shows the assistant, latest question, visitor's current page, and a link to the complete conversation.

A secure transition to Informio

The webhook is encrypted and the operator replies from the authorized administration into the original widget.

# informio-support

Live customer request

The visitor requests an operator

Informio captures the request and latest message without another AI reply.

Slack alerts the right team

The card includes context and a Take over in Informio button.

The operator continues in the same chat

The customer sees the operator's name and does not need to restart the conversation.

The first version uses Slack as a reliable notification channel. Replies stay in Informio so permissions, history, and audit remain intact.

Deployment

The entire widget is one JavaScript tag

No framework, iframe, or conflict with the customer website CSS.

  • Asynchronous loading with no dependencies
  • Server-side allowed-domain checks
  • Desktop and mobile with refresh-persistent session
Try Informio
<script
  src="https://informio.eu/widget.js"
  data-assistant="YOUR_ASSISTANT_ID"
  defer>
</script>

Turn your know-how into an assistant that answers 24/7.

Start without a payment card or API key. Default Informio AI is ready automatically, while your own OpenAI account or AI server remains optional.

FREE plan · 100 answers · no payment card