Designing a smart carpet selection system for Forohr brand

Infrastructure and site technical manager: Navid Zamani

Project manager and graphic designer: Sohail Motaghinia

Choosing a carpet for the home is not just choosing a product from several images; The carpet should be in harmony with the dimensions of the space, decoration style, furniture color, ambient light, people's taste and even the amount of daily use. This makes buying carpets online, especially for valuable and decorative products, a more complicated process than a normal purchase. In Forohar's conceptual project, Netanik with the design of a Intelligent carpet selection system It tried to turn this complexity into a simple, intuitive and personalized experience. this Intelligent carpet selection system By combining artificial intelligence, image analysis, suggestion engine and interior space simulation, it helps the user to find a carpet that is more compatible with their needs and real space. Next, pay your attention to the portfolio Khaghani carpet which is designed and implemented by Netanik.

Smart carpet selection system for the seller

The main challenge of buying carpets online

Carpet is a product whose value is not only summarized in technical specifications, and its appearance becomes meaningful along with the interior space. A design may look very attractive in a catalog image, but when placed next to furniture, flooring and wall color, it creates a different result. This gap between "seeing the product" and "imagining the product in real space" is one of the most important challenges of buying carpets online. Intelligent carpet selection system It can reduce this gap by analyzing product information and user space conditions.

The goal of the Forohar project

The goal of the project was not to build a normal online store with a few more filters, but to create an intelligent decision-making experience for the customer. The user should be able to provide a picture of his space, explain his need and finally see a selection of suitable carpets along with the reasons for the offer. The system should be able to check features such as style, color, dimensions, role and application along with interior space information. In this architecture, Intelligent carpet selection system It plays the role of a digital advisor, while the final decision remains with the customer.

The architecture of the intelligent carpet selection system

Product information layer

Any powerful bidding system needs structured and accurate data. For each carpet, information such as style, dominant color, complementary colors, dimensions, shape, material, type of texture, density, design, degree of formality or casualness and suggested use can be recorded. These data should be provided to the artificial intelligence engine in a standard form so that the system can compare different products with the user's needs. As a result, Intelligent carpet selection system It doesn't just make decisions based on the product name or a few simple labels, but uses a complete profile for each carpet.

User space recognition layer

In the second step, the system must know the user space. The user can upload the image of the room, reception or desired space and the system will extract recognizable features of the image through machine vision models. Dominant colors, the presence of furniture, the type of arrangement, the approximate level of the floor and the visual features of the space can be part of this information. Intelligent carpet selection system It then combines this data with the characteristics of the carpets to create a more tailored set of options.

General structure of the system

The proposed architecture of Forohar's system can be defined in several main layers:

This architecture makes Intelligent carpet selection system Go beyond a simple search engine and become a decision support system for purchasing. (Netanic Portfolio)

Image analysis in the intelligent carpet selection system

Knowing the color and style of decoration

One of the key capabilities of the system is visual analysis of the user's space. The vision model can be used to identify dominant colors, main objects, visual style and general features of the space. This information is then converted into data that the recommendation engine can compare with the specifications of the carpets. In this way, Intelligent carpet selection system It can extract information from the image provided by the user to personalize offers.

Identifying the important elements of space

To choose a carpet, it is not necessary to know all the details of the picture and the system should focus on the influential elements. Furniture, floor color, wall color, the amount of empty space, the overall style of arrangement and dominant colors can be prioritized in the analysis. The system can store these features with a certain level of confidence so that the user can correct the information in case of ambiguity. This method makes Intelligent carpet selection system Instead of relying entirely on AI guesswork, it has a combination of machine analysis and human verification.

Image analysis path

The image analysis process can be done in six steps:

1. Receiving the interior image by the user.

2. Recognizing the main visual elements and features.

3. Extraction of dominant color palette and complementary colors.

4. Estimation of visual style and general characteristics of decoration.

5. Receive confirmation or correction of information by the user.

6. Sending the final profile of the space to the bidding engine.

This structure makes Intelligent carpet selection system Before offering the product, first know the problem better.

Forohar's smart suggestion engine

Matching the carpet with the interior

After analyzing the image, it is time for the suggestion engine. This engine can compare the characteristics of the user's space with the characteristics recorded for each carpet and calculate the degree of matching of each product. For example, color coordination, design style, dimensions, type of use and user preference can be part of the scoring model. Intelligent carpet selection system It then presents the user with a number of options with a clear prioritization, without claiming that there is an absolutely right or wrong choice.

Personalization based on taste

The customer's taste cannot be recognized only from a picture. The user may have a minimal space but is interested in traditional carpets or, on the contrary, have a classic space and prefer a modern design. Therefore, the system should also provide the possibility to receive the direct preferences of the user and combine them with image data. As a result, Intelligent carpet selection system Instead of imposing a style, he tries to find a balance between the available space and the customer's personal taste.

Bid criteria

The recommendation engine can consider the following criteria:

• Coordinating the color of the carpet with the interior

• Match the style of the carpet with the decoration

• Suit the dimensions of the carpet with the desired space

• Direct user preference

• Space usage and usage rate

• Price range and purchase budget

These criteria can have different weights based on the type of project and be optimized over time with real user interaction data.

For more information visit Netanik's blog Please refer.

Intelligent visualization of the carpet in real space

Convert image to shopping experience

One of the most attractive parts of the Forohar project is the ability to display the carpet in the user's real space. The user can upload the image of their space and the system will generate a copy of the space with the selected carpet to get a better idea of ​​the final result. This image should not be presented as an exact simulation of execution, but should be used as a visual preview for decision making. Intelligent carpet selection system In this way, it reduces the distance between observing the product and imagining it in the real environment.

Comparing several carpets in one space

The user can choose several options and compare them in a similar environment. In this case, the difference in color, pattern, size and style of carpets will be easier to understand and decision making will be removed from the mental state. The system can also put the information of each option next to the image so that the user does not make a decision based only on appearance. The combination of visual comparison and product data is an important part of the experience Intelligent carpet selection system will be

Benefits of Smart Visualization

• Reducing ambiguity in choosing a design

• Create a better understanding of the scale of the rug

• Easier comparison of different options

• Increase user interaction with products

• Creating a more personalized shopping experience

• Helping to make decisions before buying

This feature can create more value, especially for premium carpets that take longer to make a purchase decision.

Smart assistant for buying carpets

Natural conversation with the customer

Not all users know exactly what type of carpet they want, and many of them do not express their needs in specialized language. Instead of displaying multiple filters, an intelligent assistant can start a conversation with a few simple questions and gradually determine the customer's needs. For example, the system can ask about the desired space, decoration style, favorite colors, dimensions and budget range. This assistant can be part of Intelligent carpet selection system and turn the buying process into an experience similar to a personal consultation.

Responding based on real information

The smart assistant should not generate information about the properties of carpets that are not in the product database. Answers must be generated based on actual product specifications, inventory, pricing, and verified brand information. If there is not enough information to answer, the system should refer the user to a human expert instead of guessing. This approach makes Intelligent carpet selection system Stay smart, reliable and controllable.

Shopping assistant applications

• Introduction of carpets based on decoration style

• Carpet suggestion based on interior color

• Answers to product specification questions

• Comparison of several carpet models

• Help to choose the right dimensions

• Transferring the client to a human consultant in specialized cases

Personalization of customer experience in Forohar

Build preferences profile

Every interaction of the user with the system can generate information about his taste. Viewing products, saving models, comparing options, changing filters and choosing colors can help build a preference profile. This information should be managed in compliance with privacy policies and to the extent required by the business. In this structure, Intelligent carpet selection system It can provide suggestions more tailored to the user's interest on subsequent visits.

Smart offers on the way to purchase

Personalization should not only happen on the front page of the site. The system can provide more relevant suggestions on product pages, categories, and even after viewing multiple products. For example, a user who has viewed several rugs of a certain style can receive similar or complementary products. This model makes Intelligent carpet selection system Be present throughout the purchase process and gradually personalize the user experience.

Data and SEO in Farsh online store

Product data structure

Needed carpet online store Detailed information architecture is the product. Name, image, price, inventory, specifications, brand, carpet features and related information should be placed in a structured form on the product pages. Structured data helps search engines better understand page content, and for product pages, information such as price and inventory can also be defined. As a result, data architecture is not only for Intelligent carpet selection system, but it is also important for the store's SEO infrastructure.

Carpet pages and information architecture

Categories such as modern carpets, classic carpets, Iranian carpets, hand-woven carpets, machine-made carpets, living room carpets, and bedroom carpets can be designed based on the actual needs of users. Each category should have dedicated content and understandable information, and product pages should not be just a collection of images and prices. This structure helps the search engine and the user to better understand the relationship between products and different needs. Intelligent carpet selection system It can also use the same information architecture to make more accurate suggestions.

Sales automation and customer management

Connecting AI to the store

The intelligent system becomes more valuable when it is directly connected to the infrastructure of the store. Product inventory, price, order status, specifications and customer information must be able to exchange data with the AI ​​engine within the framework of specific accesses. This connection makes offers to be generated based on the real state of the store and non-existent product or outdated information is not displayed to the customer. Intelligent carpet selection system In this architecture, it will be part of the sales infrastructure, not an independent tool and separate from the site.

Transferring the client to the expert

Not all purchasing decisions can be finalized with AI, especially when a customer has questions about specific rugs, custom orders, or high-value purchases. The system should be able to transfer the conversation to the expert at the appropriate point and provide the expert with a summary of the customer's needs. This makes the customer not have to explain his information again from the beginning. Such a hybrid architecture puts artificial intelligence and human power on a single track.

Performance analysis of the smart carpet selection system

Measurable indicators

To evaluate performance Intelligent carpet selection system Indicators should be defined that show the actual behavior of users. The amount of use of the image analysis feature, the number of interactions with the offers, the click rate on the offered products, the number of use of the visualization feature and the purchase completion rate can be part of these indicators. Each indicator should be analyzed in comparison to a specific baseline and over a suitable period of time. Without real data, it's impossible to claim that an AI system has necessarily increased sales or conversion rates.

Continuous optimization of offers

The bidding system should be continuously monitored to determine which bids users interact with the most. If users consistently reject the system's suggestions, the weighting of the criteria or the quality of the product data may need to be adjusted. Also, new products, inventory changes, and market behavior changes should be included in the bidding model. This cycle causes Intelligent carpet selection system From a fixed feature to an expandable and optimized system.

Suggested indicators

• Engagement rate with smart offers

• Rate of use of image analysis

• The rate of using carpet visualization

• Click rate on recommended products

• Compare rates of products

• Post bid completion rate

Customer data security and management

Protection of user images

Home and interior pictures can contain personal information and private environment of people. Therefore, the storage, processing and deletion of images should be done according to a specific policy and no more information than the actual system needs should be kept. It should also be clear what the images are used for and how long they will remain in the system. Data security in Intelligent carpet selection system It is part of the original architecture and should not be added after system development.

Access control and customer information

Information such as search history, preferences, images and interaction records should also be managed at the appropriate access level. The sales team may only have access to the information necessary for consulting, and the technical team may have a different need. Event logging, access control and data retention policy can reduce system operational risks. Such an approach helps Forohar's smart experience to be technically reliable in addition to AI capabilities.

The future of smart carpet selection system

Augmented reality development

The next step could be moving from visual previews to more interactive experiences. In the future, the user can view his space through the mobile camera and check different carpets in the same environment. Such a feature can bring the selection experience from a web page closer to an interactive environment. In this way, Intelligent carpet selection system It can be combined with machine vision technologies and AR experiences.

Turning the store into a smart decoration consultant

The long-term vision of the project can go beyond choosing a carpet. The system can better analyze the relationship between carpet, furniture, coloring and decoration style and become an intelligent assistant for choosing decoration elements. In this model, the customer does not just choose a product, but also receives guidance on how to coordinate it with his space. Intelligent carpet selection system At such a stage, it can become an intelligent infrastructure for the decoration shopping experience.

The result of Forohar project

The main value of the project

Forohar's project shows that AI in carpet sales can be much more than a chatbot or automatic content generator. Intelligent carpet selection system It can analyze the customer's space from the moment he enters, know his needs, suggest the right products and provide the possibility of visualizing the carpet in the environment. The combination of image analysis, product data, personalization and visual experience can reduce an important part of the ambiguity of online carpet shopping. This architecture, if implemented correctly, brings the shopping experience closer from "product search" to "intelligent consultation".

Netanik's role in the project

Netanik's role in such a project is not just to develop an artificial intelligence model, but to design an integrated ecosystem between AI, data, user experience and store infrastructure. From product profiling and suggestion engine to image analysis, intelligent assistant, product visualization and connection to the sales process, each part must be placed in a specific architecture. This approach aligns with Netanik's view of AI as part of the actual business workflow. The ultimate goal is not to build a demonstration tool; The goal is to build a system that can develop along with the growth of Forohar brand.

Finally, Intelligent carpet selection system It can bring the experience of buying a traditional and valuable product into the new generation of digital business. Instead of just searching through images and filters, the user can introduce his space to the system and receive suggestions based on the real features of the environment, his taste and needs. This approach is a combination of technology and human experience; Technology for analysis and speed, and human eye for choice and taste. For Forohar, such a system can be the starting point of a different digital experience where the authenticity of Iranian carpets and the capabilities of artificial intelligence are placed in a single path.