Challenges and Solutions: Our Expertise in Practice

Product AutoDescriber AI – Effortlessly generate product descriptions using advanced language models​

Introducing Product AutoDescriber AI, an innovative module for the Odoo platform that employs the power of OpenAI's ChatGPT to automatically generate distinctive and engaging product descriptions from photographs. In the realm of modern e-commerce, unique and high-quality product descriptions are essential for increasing sales and optimizing SEO. The process of manually writing descriptions for a large inventory can be both time-consuming and inefficient.

To solve this problem, Wolfinne IT has developed Product AutoDescriber AI, an Odoo module that uses OpenAI GPT technology to automatically generate product descriptions based on their images. This tool enables fast, accurate, and contextually relevant creation of descriptions, reducing manual work and optimizing content quality.

Product AutoDescriber AI is available on the Odoo apps market at the link: https://apps.odoo.com/apps/modules/17.0/prod_desc_gpt

Key functionalities:

• Automatic generation of product descriptions: Using the OpenAI GPT model and OpenAI API, the module analyzes product images and generates descriptions tailored to the given tone. It is possible to configure the version of the used GPT model.​
• Support for multiple languages and customized communication tone: The system allows for the generation of content in different languages depending on which language is selected for the product, with the option to customize the tone of the description (formal, humorous, etc.).
• Seamless integration with Odoo e-commerce and product catalog: The module easily connects with Odoo e-commerce and CRM system, allowing for direct creation and editing of descriptions within the Odoo environment. It is possible to generate descriptions for individual products or for all products in the database.
• SEO optimization and better visibility in searches: Automatically generated descriptions are SEO-friendly, improving product visibility in searches and increasing organic traffic to e-commerce.​
• Quick implementation and easy setup of the API key: After activation, you only need to enter the OpenAI API key and the system is ready to go.

Challenges and solutions:

1. Challenge: Creating high-quality product descriptions at scale
   • Problem: E-commerce stores with a large number of items often do not have enough resources to manually create unique descriptions.​
   • Solution: The AI model enables fast and automatic text generation, tailored to the visual characteristics of the product.​
2. Challenge: Precision and Context in Generated Descriptions
   • Problem: AI tools often create generic descriptions that do not reflect the specific characteristics of the product.​
   • Solution: We have implemented fine-tuning mechanisms and advanced filters to ensure that the generated texts are as accurate and relevant as possible for the target audience.​
3. Challenge: Integration with the Odoo e-commerce system
   • Problem: Product descriptions must be directly available in the Odoo catalog and easily editable within the ERP system.​
   • Solution: The module was developed as a native Odoo application, allowing for automatic connection to existing data in the product database.​
4. Challenge: Increasing conversions through better descriptions
   • Problem: The generated descriptions must be not only informative but also sales-oriented.​
   • Solution: The model has been trained to create marketing-optimized descriptions that highlight the key benefits of the product and enhance the user experience.​

🔹 Technologies: Odoo · PostgreSQL · Python · JavaScript​

Metalpine – Integrated business solution based on Odoo ERP

The Wolfinne IT team developed a comprehensive business solution for Metalpine, an innovator in the field of sustainable metal powders, using Odoo Community Edition and its extensive capabilities. The project included complete integration of modules for accounting, human resources, manufacturing (MRP), and CRM, creating a unique and efficient system tailored to the specific operational needs of the company.​

Our team was responsible for system design, module customization and implementation, infrastructure, and operations (DevOps), ensuring seamless functionality across all business segments. This customized Odoo system significantly improved operational efficiency, data connectivity, and business analytics capabilities, providing Metalpine with a solid foundation for further growth and scaling of the business.

Wolfinne IT has developed a complete business solution based on Odoo ERP for Metalpine, a leader in the production of sustainable metal powders, optimizing key processes from sales and production to human resources and analytics.

This integration enables a transparent and efficient business flow, connecting different sectors into a unified digital system. The application includes:

• Sales module – Enables order tracking, invoicing, and analysis of sales performance, along with automated document creation and integration with customers.
• Inventory and Production Management – Through the MRP module, Metalpine precisely tracks materials, production processes, and optimizes costs, achieving better efficiency in the supply chain.
• Human resources and working hours – The system enables the management of shifts, working hours, and employee records, ensuring efficient organization of teams and resources.
• Business analytics and dashboard views – Management has an overview of key indicators in real-time, including sales metrics, billed amounts, team performance, and optimization of work processes.

By implementing this modular ERP system, Metalpine has gained a centralized platform for managing its entire business, ensuring greater productivity, better data control, and optimized resources..

Challenges:

1. Integration of different business processes into a unified ERP system​
   • Metalpine has used various tools and systems for sales, production, inventory management, and human resources.
   • It was necessary to consolidate all processes into a unified ERP platform to ensure data consistency and more efficient operations.
2. Precise monitoring of production and resource optimization
   • The production of metal powders involves a complex supply chain, quality control, and material optimization.
   • The ERP system had to enable automatic management of inventory and resources, minimizing human errors.
3. Sales and Invoicing Automation
   • Manual tracking of orders and invoices hindered the efficiency of business operations.
   • The system had to enable automated invoice generation, notification sending, and integration with customers.
4. Customization of the user interface and employee training
   • ERP solutions often come with a complex user interface, which can make it difficult to adopt the new system.
   • Customized training for employees and optimization of UI/UX design were necessary to enable easy navigation through the system.

Solutions:

1. Unique Odoo ERP platform for all business processes
   • By integrating the sales, MRP, inventory, and human resources modules, a centralized system has been created that connects all sectors of the company.
   • Automatic data synchronization has enabled the accuracy and consistency of information in real time.
2. Production optimization using the MRP module
   • A MRP (Manufacturing Resource Planning) system has been implemented that enables automatic management of inventory, materials, and production processes.
   • Real-time inventory management helps reduce waste and optimize production costs.
3. Automated invoicing and sales management
   • A sales module has been developed with capabilities for automatic invoice creation, sending notifications to customers, and analyzing sales performance.
   • Sales analytics has been implemented, allowing management better insight into results and trends.
4. Customized user interface and employee training
   • Intuitive dashboard views have been created, allowing for easier navigation through key business indicators.
   • Employee training was conducted, which increased the efficiency of using the ERP system and reduced the need for additional IT support.

Through the integration of Odoo ERP, Metalpine has acquired a fully digitized business system that connects sales, production, logistics, and human resources into one platform. This solution ensures greater efficiency, better control over resources, and cost optimization, making the company ready for further growth and scaling of its operations.

🔹 Technologies: Odoo 13 Community Edition · PostgreSQL · Python · JavaScript

GRAWE Austria – Web application for tracking insurance claim status

Wolfinne IT has developed a mobile-optimized web application for GRAWE Austria, allowing users to track the status of their vehicle damage claims in real time. This system significantly enhances customer support and operational efficiency, providing policyholders with transparent insight into the claims processing procedure.

The application was developed using Angular for a dynamic and intuitive user experience, while the Django REST Framework provided a stable and scalable backend. A key functionality of the system is the advanced integration with GRAWE's internal K3 system, enabling automatic notifications to users via email and SMS about the status of their requests. This solution has established an efficient and transparent communication channel, ensuring that clients have accurate information about the claims processing at all times.

This solution significantly increases the efficiency and transparency of insurance processing, reduces the administrative burden on GRAWE customer support, and improves customer satisfaction through a modernized digital approach to insurance tracking.

Challenges:

1. Secure and easy access for users
   • Users needed to be provided with a fast, yet secure way to log in, without the need to create an account and remember passwords.
   • The solution had to meet strict data protection standards, especially since it involved sensitive information about insurance and claims.
2. Transparency and automation of communication
   • Policyholders previously had to contact support by phone to obtain information about the status of their claim.
   • It was necessary to develop a system that automatically notifies users of status changes, reducing the burden on customer service.
3. Integration with the internal GRAWE K3 system
   • The request tracking system had to be fully aligned with GRAWE's internal processes and enable seamless data exchange.
   • Real-time data synchronization required optimized integration with the existing IT infrastructure of the insurance company.
4. Mobile adaptability and intuitive user experience
   • The application had to be easily accessible on various devices, especially mobile phones, so that policyholders could quickly check the status of their application.
   • The design had to be clean, simple, and intuitive so that users of all generations could easily use the application.

Solutions:

1. Login using a one-time password (OTP) for security and simplicityt
   • A login system has been implemented using a six-digit password (OTP) that users receive via SMS or email.
   • Time provides fast and secure access without the need to create an account and complicated passwords.
2. Automatic notifications via email and SMS
   • Users can choose their preferred communication channel (email or SMS) to receive updates on the status of their request.
   • Automated notifications reduce the need for manual tracking and contacting customer support.
3. Integration with the K3 system for real-time data monitoring
   • The application is connected to GRAWE's internal K3 system, allowing the status of applications to be automatically updated and displayed in the application.
   • Fast data exchange is ensured with high security standards and optimized performance.
4. Mobile optimization and user-friendly design
   • The application is designed with a "mobile-first" approach, ensuring a seamless experience across all devices.
   • The user interface has been simplified, with clear information and intuitive navigation, allowing users to quickly review all key information about the status of their requests.

Through carefully designed solutions, the application has significantly improved the user experience, accelerated the processing of requests, and reduced the administrative burden for the GRAWE team. The digitalization of this process has brought greater efficiency, transparency, and simplicity, making insurance more accessible and modern for all users.

🔹 Technologies: Angular · Django REST Framework · MongoDB

Platform Against Loneliness plattform-gegen-einsamkeit.at

Platform Against Loneliness is an initiative launched in 2021 by Social City Wien with the support of the Austrian Ministry of Social Affairs, aimed at providing resources and information to combat loneliness and social isolation in Austria. The project was originally developed on the WordPress platform, but due to the need for modernization and better integration of functionalities, it was migrated to the Odoo Website solution.

What has been done:

• Content migration: All existing content, including pages, blogs, media files, and user data, has been transferred from WordPress to the Odoo platform.
• User Interface Redesign: The capabilities of the Odoo Website Builder have been utilized to create a modern, responsive design that provides a better user experience across different devices.
• Integration of functionalities: Additional features have been implemented such as interactive maps for searching regional offers, contact forms, and event registration, as well as advanced storage and contact search functions.

Challenges and solutions:

1. Challenge: Transferring the entire database and media content from WordPress to Odoo without data loss​.
   Solution: A custom script has been developed that uses the WordPress REST API for data extraction, which is then transformed and imported into Odoo using its built-in data import tools​.
2. Challenge: Maintaining SEO rankings and existing URL structures to avoid losses in search traffic.
   Solution: Custom URLs have been configured in Odoo to match SEO requirements.
3. Challenge: Team training for using the new platform and ensuring continuity in content updates.
   Solution: Workshops were organized that enabled the team to quickly adapt to the new system and efficiently manage content.

By migrating to Odoo Website, the Plattform gegen Einsamkeit has gained a modern, flexible, and scalable platform that better meets the needs of users and allows for easier administration and further growth of the platform.

🔹 Technologies: Odoo 18 · JavaScript · SVG · PostgreSQL · WordPress API

Prototype AI tool for preventive maintenance of high-voltage electrical networks

High-voltage networks require continuous monitoring and timely detection of potential faults to ensure the reliability and stability of the power system. Traditional inspection methods, which rely on manual checks with field teams, are often expensive, time-consuming, and inefficient, creating a need for innovative technological solutions based on artificial intelligence (AI) and machine learning (ML).

This AI tool for preventive maintenance of high-voltage networks is based on the analysis of data collected by drones, which enable precise aerial imaging of infrastructure. The imaging includes LIDAR data, RGB orthophoto images, close-range and thermal photographs, providing a detailed view of the network and its surroundings.

Wolfinne IT provided consulting and development services on this project, actively participating in the creation of a prototype AI tool for analyzing and predicting failures, which is shown in the video below. In addition to developing the software solution, our team was involved in developing a data acquisition methodology, creating an optimized flight plan for drones to ensure efficient and standardized coverage of high-voltage lines.

By combining automated recording, data analysis using AI models, and predictive analytics, earlier detection of problems has been enabled, thereby reducing maintenance costs, extending the lifespan of infrastructure, and improving the safety of electrical distribution networks.

What has been done:

• System architecture design: We created a complete architecture for the software solution, connecting data processing from multiple sources (LIDAR, RGB, thermal images) with an analytical layer based on AI models.
• Development of the backend and prototype frontend: We implemented a backend system that enables the processing, storage, and analysis of data, as well as a prototype frontend application for user review and interaction with the data.
• Integration with AI models: Our AI collaborator has developed ML models for data analysis, and we have integrated an ML infrastructure based on MLRun, enabling scalable and efficient processing of large datasets.
• Analysis of images from multiple sources: The AI model analyzes LIDAR images, as well as RGB orthophotos, close-range and thermal photographs, detecting potential failure points and providing predictions based on historical data and current changes in the network.

Challenges and solutions:

1. Processing large amounts of data from various sources
   • Challenge: LIDAR, orthophotos, and thermal images generate large datasets that must be efficiently stored and analyzed.
   • Solution: We have implemented optimized methods for data storage and processing using PostGIS for georeferenced data and distributed processing to accelerate AI analysis.
2. Precise detection of potential network faults
   • Challenge: AI models must recognize early signs of failures so that the system can suggest preventive actions.
   • Solution: Using deep learning and models trained on previous data, a high accuracy in anomaly detection in the network has been achieved.
3. Visualization of analysis results in an intuitive way
   • Challenge: Presentation of AI analysis and results to users in an easily understandable format.
   • Solution: We have developed an interactive interface that allows for data visualization on maps, comparison of historical snapshots, and generation of reports for technical teams.
     

This AI tool for preventive maintenance of high-voltage electrical networks brings a revolutionary approach to the maintenance of electrical power systems. Through automated analysis and predictive problem detection, it enables a reduction in failures, cost optimization, and an increase in network reliability.

Wolfinne IT has contributed to the creation of an innovative, scalable, and efficient system that enhances the security and sustainability of electrical distribution networks through architecture design, backend and frontend development, and integration with AI models.

🔹 Technologies: Django REST Framework · React · MLRun · PostgreSQL · PostGIS · MongoDB · Redis · AWS 

Platform for calculating the profitability of investments in solar power plants - www.SolarInvest.ba  

www.SolarInvest.ba is an interactive platform dedicated to the promotion and development of solar energy in Bosnia and Herzegovina. Its goal is to enable users to assess the profitability of solar investments, provide comprehensive information about the country's solar potential, and connect them with relevant experts and institutions. The project was led and funded by UNDP.

As part of providing consulting services and technical expertise, Wolfinne IT played a key role in leading a multifunctional development team, ensuring effective coordination between backend and frontend developers, GIS specialists, and data analysts.

Our participation in the project included:

• Technical consultations and management of the development team: We organized development processes, aligning the team's work through agile methodologies. We defined technical standards and a development strategy to ensure that the platform is scalable, reliable, and sustainable in the long term.
• Product and backend system development: We actively participated in the design and implementation of the platform's key functionalities. We implemented a robust backend system that enables efficient processing of large amounts of data and provides stable APIs for the frontend and external services.
• Integration with external data and precise analysis of solar potential: We have developed algorithms that use GIS data and meteorological information to simulate solar irradiance and energy production. We have optimized processes for handling large datasets to provide users with fast and accurate analyses.

Challenges we faced and our solutions:

1. Coordination and efficient work of a multifunctional development team
   • Challenge: Organizacija rada između različitih specijalizovanih timova (backend, frontend, GIS, analitičari podataka).
   • Solution: Uveli smo agilne metodologije i redovne sprint review sesije kako bismo osigurali efikasnu komunikaciju i sinhronizaciju rada svih članova tima.
2. Integration of different data sources in real time
   • Challenge: Combining GIS data, meteorological information, and energy capacity data into a unified platform.
   • Solution: We have developed advanced ETL processes that automatically collect, process, and synchronize data, ensuring their consistency and accuracy.
3. Precise modeling of solar potential and simulation of energy production
   • Challenge: Enable users to have reliable simulations based on geographic locations and weather data.
   • Solution: We have implemented algorithms that analyze solar potential based on historical data and two different meteorological data models.
4. Performance optimization and system scalability
   • Challenge: Enable the platform to process large amounts of data and support a high number of users without losing performance.
   • Solution: We used a microservices architecture with horizontal scaling and optimized databases to ensure fast system response even under high load.
     

As a result of our engagement, www.SolarInvest.ba has become a reliable and technologically advanced tool for anyone looking to invest in solar energy. We provided technical expertise and led the development team, ensuring that the platform is modern, scalable, and functionally rich. This project represents an important step towards energy sustainability and the digitalization of solar investments in Bosnia and Herzegovina.

🔹 Technologies: Django REST Framework · React · PostgreSQL · PostGIS · GeoServer · AWS

SolarVision AI prototype - a tool for determining the optimal area for photovoltaic systems

SolarVision AI is an advanced prototype tool designed for identifying optimal areas for the installation of photovoltaic (PV) systems through the analysis of satellite imagery. By combining artificial intelligence and machine learning technologies, SolarVision AI enables precise segmentation and evaluation of terrain, facilitating the planning and implementation of solar energy projects.​

Key features of the SolarVision AI prototype:

• Analysis of satellite images: By using high-resolution satellite data, the tool analyzes the geographical characteristics of the area of interest in detail.
• Terrain segmentation: By applying advanced image processing algorithms, SolarVision AI identifies different types of surfaces, such as rooftops, open land, and bodies of water, determining their potential for installing PV systems.
• Evaluation of Sunlight Exposure: The tool estimates the level of solar irradiance for each identified surface, taking into account factors such as geographical latitude, terrain slope, and local climatic conditions.
• Installation Recommendations: Based on the collected data, SolarVision AI generates recommendations for optimal locations and orientations of PV panels, maximizing energy efficiency and the profitability of the investment.

In the video shown below, the functionality of the SolarVision AI prototype is demonstrated through concrete examples of satellite image analysis and the identification of the most suitable areas for the installation of solar panels. The video illustrates how the tool automatically recognizes and classifies different types of terrain, assesses their solar potential, and provides the user with visual and numerical information to make informed decisions about the installation of photovoltaic systems.

The implementation of the SolarVision AI prototype represents a significant step towards more efficient use of renewable energy sources, enabling investors and planners to quickly and accurately identify the most suitable locations for solar projects, thereby reducing costs and the time required for planning and execution. From this prototype, the idea of a comprehensive tool for designing photovoltaic systems emerged, which is described in the document available at this link:

Challenges and solutions:

1. Precise segmentation of satellite images
   • Challenge: Satellite images often contain noise, shadows, and low resolutions, making it difficult to accurately identify areas suitable for PV systems.
   • Solution: An advanced segmentation model has been developed that uses CNN (Convolutional Neural Networks) and optimized filters for recognizing and categorizing surfaces.
2. Data Integration from Different Sources
   • Challenge: Data on buildings, land, and solar irradiation come from various sources and in different formats, making their unification difficult.
   • Solution: We have implemented an ETL (Extract, Transform, Load) process that automatically collects, standardizes, and synchronizes data to ensure the accuracy of estimates.
3. Accurate assessment of solar potential
   • Challenge: It is necessary to take into account the slope, orientation, climatic conditions, and annual variations in solar radiation.
   • Solution: The tool uses historical climate data in combination with GIS analysis and AI models to generate accurate energy simulations.

🔹 Technologies: Django REST Framework · React · PostgreSQL · PostGIS · MLRun · AWS