A solution for e-commerce

JediDesk for online stores

Help every shopper find what suits them.

JediDesk helps online stores advise shoppers when there are more questions than your team has time for. AI clarifies their needs, selects products from your catalogue and answers according to your store's rules. A connected order system provides data for support after purchase, while complex enquiries go to your team along with the conversation history.

Try for freeExplore the scenarios

Catalogue · knowledge base · team

Art collage: Botticelli's Venus choosing cosmetics on a laptop, beside a symbol of digital conversation
Attention to every question
Beauty meets technology. An AI interpretation of Botticelli's Venus.
For store owners and support teamsUpdated

Your store's day-to-day work

Shoppers need answers when they are choosing

A product page lists specifications, but every shopper has their own situation. They need to understand what suits them, how the options differ and how to receive their purchase. These are the conversations that recur every day.

Choosing between similar products

“How do these two products differ?” or “What would make a good gift?” — shoppers expect an explanation that takes their needs into account.

Enquiries from different channels

Shoppers write through website chat, Instagram and Telegram. When an adviser joins the conversation, they need to see the earlier messages within that enquiry.

Questions after payment

Order status, delivery, payment and returns need specific answers. Some information is in the store's rules; some is only in its business system.

Peaks during promotions

After a campaign message or an ad launch, questions about products and purchase terms arrive at the same time. Your team balances advice with order processing.

Before automation

What takes up your store team's time

The same explanations again and again

Your team repeatedly explains product differences, payment methods and delivery terms. Less attention remains for more unusual product requests.

The shopper's need is still unclear

“Could you recommend something?” starts a series of follow-up questions. Without them, the recommendation becomes a list of links the customer has to review alone.

Looking for data in different places

To answer, your team has to open the product page, store rules and order system. These are different sources, each of which must be kept up to date.

Handovers without details

If an enquiry passes to a colleague without an explanation, the shopper repeats their needs, product code or delivery question. Conversation history helps your team continue from where it left off.

Automation use cases

From a question to a clear next step

Configure the assistant around your store's products and rules. Every scenario has an information source and a clear point for handing over to a person.

  1. 01Shopper's needs
  2. 02Catalogue data
  3. 03An informed choice
Art collage: Girl with a Pearl Earring comparing cosmetics using a smartphone
Choosing a product — a modern scene with Girl with a Pearl Earring. An AI interpretation.
01

Selection based on need

The shopper describes what they need but does not yet know the product name.

AI clarifies the important criteria and suggests options from the connected catalogue, explaining why they were selected.

Product selection with JediDesk AI

The shopper receives a selection they can use to continue choosing.

02

Comparing options

There are several similar items in the basket or saved list.

The assistant explains differences using specifications available in your product data. If a specification is missing, it checks with your team.

Guidance through website chat

The customer understands which differences matter for their purchase.

03

Availability, payment and delivery

Before placing an order, the shopper needs purchase terms and product information.

AI uses the store's catalogue and knowledge base. How current stock levels are depends on the connected source and its updates.

A knowledge base for AI

Routine questions are handled in the same conversation.

04

Order status

The shopper asks what stage their order has reached.

With an integration configured, the assistant retrieves order data. If there is not enough information to find the order, it asks the customer for clarification.

E-commerce integrations

The answer is based on the store's business system.

05

Complex enquiries go to your team

An unusual request or a problem with a purchase needs a decision.

AI gathers details and passes the enquiry to an operator according to your rules. The team member sees the conversation history and continues it.

AI and operators working together

A person joins with the situation already explained.

How the solution works

Your data. Consistent guidance.

The assistant works with information your store has prepared and connected. The catalogue, purchase rules and order data play different roles in the conversation.

  1. 01

    The shopper gets in touch

    They write in website chat or a connected messaging channel, describing a need, naming a product or asking about an order.

  2. 02

    AI clarifies the task

    It gathers the selection criteria or details needed to find an order. If information is missing, the configured scenario provides for clarification or a handover to your team.

  3. 03

    It consults the source

    It uses the catalogue for selection, the knowledge base for purchase terms and a configured integration for order status.

  4. 04

    It helps or hands over

    It provides an answer and a next step. Questions outside the rules go to your team, which reviews conversations in JediDesk.

Product catalogueStore rulesOrder dataContext for your team

Capability map

What your store's AI can do

Every capability relies on your data and rules. Here is how they come together around one assistant.

A shopper's enquiry in chat or a messaging channel

Botticelli's Venus in a blue headset at a desk with a laptop and cosmetics — an artistic depiction of an AI assistant
JediDesk AIAn assistant for shoppers and your team

  • Selection tailored to the enquiry

    Clarifies the need and suggests products, explaining the selection.

    Source: Catalogue + shopper criteria

  • Price and availability

    Uses connected data. Its accuracy depends on updates.

    Source: Product feed / integration

  • Details and comparisons

    Explains features and differences using product descriptions.

    Source: Product specifications

  • Order status

    Checks the order stage with an integration configured.

    Source: Order system

  • Payment and delivery

    Explains payment methods, delivery and return terms.

    Source: Store knowledge base

  • Handover to your team

    Passes on a complex enquiry with its conversation history.

    Source: Handover rules + team

The shopper receives an answer or help from your team with the conversation context.

An AI interpretation of Venus. An illustrative capability map.

Channels and integrations

Meet shoppers where they message you

Choose the channels you need and connect your data sources. The capabilities of each integration are agreed around your store's process.

Website chat

Help alongside the product page when a shopper compares options.

About the channel

Instagram

Guidance on products the shopper has seen on your profile.

About the channel

Telegram

A conversation with the shopper through your store's connected bot.

About the channel

Viber

Enquiries to your store through a messaging app familiar to the customer.

About the channel

Catalogue and order system

Horoshop and other e-commerce integrations: agree which products, fields and order data the assistant can access.

How to connect

Knowledge base and AI rules

Prepare payment, delivery and return terms, and instructions for passing enquiries to your team.

How to connect
Art collage: a marble Hermes wearing a support headset beside a laptop and parcels
Hermes as a modern order assistant. The illustration was created using AI.

Example conversations

What it sounds like in a conversation with a shopper

Illustrative store scenarios. These show the logic of the consultation; they are not real customer conversations or demonstrations of specific product recommendations.

  1. 01Enquiry
  2. 02Order details
  3. 03Help from your team

Help with choosing

Website chat
  1. Shopper

    I'm looking for a moisturiser. You have so many — where should I start?

  2. JediDesk AI

    Please tell me which format you prefer and what matters to you when using it. I'll select options from the store's catalogue.

  3. Shopper

    A light cream without a strong fragrance.

  4. JediDesk AI

    I'll check the product descriptions and specifications. If the catalogue does not state fragrance intensity, I'll check with an adviser.

A handover to an adviser if the required specification is missing from the source.

Selection starts with the shopper's needs and product data.

Hand over a problem with context

Telegram
  1. Shopper

    I received my order, but one item is damaged.

  2. JediDesk AI

    Please send the order number and describe which item is damaged. I'll gather the information for our team.

  3. Shopper

    I've already sent the number. I can add a photo of the packaging.

  4. JediDesk AI

    Please add the photo to this conversation. I'm passing your enquiry to our team so they can check the details and agree the next steps with you.

Your team receives the enquiry and conversation history.

The decision on a replacement or return follows the store's rules.

Results and benefits

What the shopper and your team gain

The effect depends on your product range, data quality and configuration. Start with scenarios your team already knows well and check the results in your own conversations.

Selection with an explanation

The shopper sees how their needs relate to the suggested product features.

Fewer repeated explanations

Prepared payment and delivery rules are used for routine questions in conversations.

Context for your team

A person receives the situation description and previous messages when an enquiry needs their involvement.

Information to guide improvements

Conversations reveal which specifications are missing from the catalogue and which rules need clearer explanations.

Customer experience

How businesses use JediDesk

About TANOYA

Cosmetics are chosen around a need. The consultation works the same way.

An illustrative visual from the TANOYA case study. The process description is based on the case study materials.

In the TANOYA case study, JediDesk AI helps cosmetics shoppers clarify their needs and select products from the catalogue. Other scenarios cover basic purchase questions and handing complex enquiries to the team.

  • Selection is based on the shopper's request and product information.
  • Advice and routine support are combined in one process.

Built around your process

Why JediDesk suits online stores

Working with your product range

Recommendations are based on connected product data. Your store decides which information the assistant uses and what needs clarification.

JediDesk AI capabilities

Connections to your business systems

Scenarios can be supplemented with order data through an integration. Available actions depend on your store's system and its connection.

JediDesk integrations

AI works alongside your team

An operator joins the conversation when a human decision is needed. The knowledge base and handover rules help define the assistant's responsibilities.

AI assistant setup

Launching the solution

Start with your store's scenarios

Launch needs someone who knows your products and support rules, and someone responsible for the website or integration. The timeline depends on data readiness and connection complexity.

  1. 1

    Prepare your data

    Gather your catalogue or feed, product specifications, common questions, and payment, delivery and return rules.

  2. 2

    Connect your channels

    Create an account and add chat or a messaging channel. Work with a technical specialist to configure the sources and integrations you need.

  3. 3

    Set the boundaries of guidance

    Define what the assistant can explain itself, which details it clarifies and when it passes an enquiry to a specific team.

  4. 4

    Test and improve

    Walk through routine and complex scenarios. After launch, review conversations and expand the catalogue and knowledge base.

JediDesk AI setup guide

Before you start

Frequently asked questions

Answers about the catalogue, orders and AI working with your team.

Can AI select products specifically from our catalogue?

Yes. The selection scenario is configured using your store's product database or feed. The assistant uses available names, specifications and other connected fields to match a shopper's needs to suitable items. Before launch, check data completeness and update rules. If an important specification is missing, plan for clarification through your team rather than an AI assumption.

Will the assistant be able to report order status?

This scenario requires an integration with the system that stores orders. Define which details are used to find a purchase and which fields can be returned to the customer. A product feed alone is not enough. If an order is not found or information is missing, the assistant should clarify the details or pass the question to your team according to the store's configured process.

How do we keep prices and availability up to date?

The assistant relies on the connected source, so answers are only as current as the store's updates to prices, stock and product range in that source. During integration, agree the update method and frequency, and behaviour when data is unavailable. Before launch, test scenarios with a changed price, unavailable product and new item. Availability promises must match the capabilities of the specific connection.

When does a team member join the conversation?

Your store sets rules for handing an enquiry to a person, for example purchase complaints, unusual orders or questions outside the knowledge base. AI can gather a situation description and the details needed before handing over. Your team continues with the conversation history. Decisions on refunds, product replacements or other exceptions remain within the authority defined by your process.

Do we have to connect every messaging channel at once?

No. You can start with the channel where your store receives the most routine questions, such as website chat or Telegram. First check answer quality, catalogue selection and operator handovers. Then add other channels you need and test the scenarios in each. Available capabilities and connection requirements depend on the particular channel and account settings.

What should we prepare before launching JediDesk?

You need product data, clear purchase rules, examples of routine enquiries and a list of questions for your team. Assign responsibility for updating the catalogue and knowledge base. If you plan to check orders, prepare a technical description of the business system and the access needed for integration. Before working with shoppers, test answers using sample scenarios; the launch timeline depends on the readiness of these materials.

Other industries

Find your next use case

The next step — your store

Give shoppers help when they are choosing

Connect your store's knowledge and start with the most common questions. Test JediDesk with your own products and scenarios.

Try for freeSign up

The free trial lasts 14 days. Integration setup depends on your system.

  • 14 days to explore
  • Your products and rules
  • AI and your team in one process
Top