About the client
Services
Expertise delivered
Technologies
EXECUTIVE SUMMARY
About Gepard
Gepard is a PIM platform that brings innovative automation solutions into product information management for eCommerce businesses. Gepard PIM & Syndication platform is a single source of truth to collect, manage, enrich, and distribute users’ product data in the required format to various sales channels. It enables brands to exchange product marketing content freely and helps retailers to onboard and adapt it in an effective and automated way.
Their solution increases operational efficiency by 75% and delivers 120+ million product descriptions per month across multiple retail platforms.
Services Provided
Gepard has pioneered in offering cutting-edge AI Mapping solutions as part of its comprehensive Product Information Management platform. Our services focus on automating and streamlining the process of category mapping, feature mapping, and feature value mapping, alongside an innovative table-based import system for eCommerce businesses.
The AI Assistant integration unlocks breakthrough platform features:
- automatic recognition of data import file fields;
- automatic mapping of feature values;
- providing feature value suggestions;
- AI-generated category and feature mapping suggestions;
- AI-assisted shortcuts for one-click mappings.
Advanced benefits for the users:
- ready-recognized fields of data import files for faster taxonomy mapping;
- less time and manual work for the mapping to find relevant values in the content taxonomy;
- next to zero time and manual work to map features, feature values, and categories of the content and
- external taxonomies that completely match.
Technologies Used
Our AI Mapping development leveraged advanced technologies including the integration of the OpenAI API embedded model for intelligent category, feature, and feature values recognition, and OpenAI API algorithms for supporting new taxonomies and enhancing mapping accuracy over time.
Team Composition
The project has been to life by a dedicated team comprising:
- 1 solutions architect;
- 2 Backend and 1 Frontend for the code development and UI integration;
- 1 AI specialist focusing on the OpenAI API embedded model integration;
- 1 project manager overseeing the project timeline and deliverables;
- 1 QA engineer ensuring the quality, reliability, and user-friendly applicability of the AI-powered
- mapping.
Challenges Bintime Team Addressed
We faced several challenges, including:
- The complexity of managing vast amounts of product data and displaying the result on the UI within the time acceptable for users;
- The need for a flexible mapping system that could handle simple mappings, keeping in mind further AI-powered suggestions for complex mappings;
- Ensuring the system’s reliability and scalability to support the processing of new taxonomies and products without any additional manual efforts.
Our Approach
The project embarked on several pivotal stages, beginning with a comprehensive Discovery Phase aimed at grasping the distinct requirements and hurdles encountered within eCommerce operations. Following this, the team delved into engineering the architecture solution tailored to seamlessly integrate into the platform infrastructure.
Subsequently, efforts were directed towards the meticulous integration of the OpenAI API embedded model, facilitating the implementation of an intelligent mapping system characterized by a responsive user interface. Rigorous QA/QC testing procedures were then undertaken to guarantee the reliability and precision of the mappings, ensuring a seamless user experience. Furthermore, a commitment to continuous updates and support was upheld, enabling the adaptation to evolving market dynamics and incorporating valuable feedback to enhance the overall functionality and efficiency of the system.
Our strategy encompassed a blend of methodologies to ensure agility and innovation:
- Agile development for adaptive planning and continuous improvement of our MVP features;
- Component-Based Approach to ensure modularity and ease of maintenance;
- CI/CD development for streamlined and automated code deployment;
- Focus on user experience to make the platform intuitive and easy to use;
- Security-minded design to protect data integrity and privacy.
Value Delivered
Gepard’s AI Mapping development has resulted in:
- The accuracy of auto-mapping suggestions hit the mark of 75%, enabling brands and retailers to manage product information more effectively.
- Delivering accurate and automated category, feature, and feature value mappings, reducing manual effort and errors.
- Enhancing scalability to support the mapping of new taxonomies and products across various retail platforms.
- Providing an intuitive user interface for easy navigation and operation of the mapping functions.
Current State of Collaboration
The OpenAI API embedded model has been integrated into Gepard’s PIM & Syndication platform to cover the data mapping — the most complex process of product content syndication and enrichment.
Gepard’s AI Mapping is a testament to our commitment to innovation and excellence in product information management. By harnessing the power of AI, we have set a new standard in the eCommerce industry, enabling brands and retailers to achieve unprecedented levels of efficiency and accuracy in managing their product data.
linkedin