An AI chatbot that answers questions related to the CNC industry.
Their product, Gieni, is the world's first chat-based tool for B2B and industry. It answers specific questions about the CNC manufacturing industry and receives deep market insights from the specialized database. These data help to strategically expand market share, analyze competition, optimize sales, launch new products and successfully steer innovations in the CNC industry.
NestJs
Typescript
RestAPI
React
React-map-gl
Jest
TanStack Query
SCSS
react-i18next lib
Commitizen
Stripe
React Testing Library
Recharts lib
Client’s brief
- Acquire a comprehensive, reliable, and relevant global database of CNC manufacturers
- Train an AI model to understand CNC-related queries
- Ensure data security and ethical use of AI
- Create intuitive UX/UI
- Implement multilingualism
- Implement Natural Language Understanding (NLU)
- Offer the most effective monetization and user engagement model
Our Part
Initiating an AI chatbot can encounter various challenges at the outset. Here are some challenges we faced while working on the MVP version of the product and the solutions we implemented.
1. Data Collection and Quality
Data Sourcing
We assembled a comprehensive and reliable global database of CNC manufacturers' machines, capabilities, and materials. That was challenging due to fragmented and disparate sources of information.
Data Quality
We have verified the data's accuracy, relevancy, and currency. Inaccurate or outdated information can affect the chatbot's responses and user experience.
2. AI Model Training and Accuracy
Training Data
We curated and prepared a diverse dataset for training the AI model to understand CNC-related queries accurately. Insufficient or biased training data can lead to inaccurate responses.
Precision and Recall
We achieved a balance between precision (accuracy of relevant responses) and recall (coverage of all relevant information). Such a task is a common challenge in AI-driven systems.
3. User Experience and Interface Design
User Experience and Interface Design
We designed a user-friendly dashboard that presents maps, charts, diagrams, text, and other content in a coherent and understandable format.
Intuitive Interaction
We were guided by an intuitive design approach so that users can easily navigate the chatbot, understand its capabilities, and access desired information without friction.
4. Natural Language Understanding (NLU)
Domain-Specific Language
CNC-related terminology and industry-specific jargon can be intricate and diverse. We taught the AI to comprehend and interpret these nuances accurately.
Ambiguity Handling
We resolved ambiguities in user queries to provide precise responses. It is crucial for AI-based solutions, especially in technical domains where context plays a crucial role.
5. Monetization Model and User Engagement
Balancing Free and Paid Access
We balanced free and paid access to entice users while maintaining revenue.
Feedback Loop Implementation
We implemented a solution to encourage user feedback and integrate it into the system effectively to improve the chatbot's performance and offer more free questions.
6. Regulatory and Ethical Considerations
Data Privacy
We developed our solution in compliance with data privacy regulations and ensuring the security of user information.
Ethical Use of AI
We ensured AI's responsible and ethical use, especially while providing market insights or recommendations.
7. Technical Infrastructure and Integration
Database Integration
We solved the technical challenge of integrating and synchronizing a global database into the AI system while maintaining its performance and accuracy.
Scalability
We designed the system with the ability to handle increasing user demands and scaling infrastructure to accommodate growing user bases.
Solutions for the MVP version
User Authentication
Seamless login functionality for users to access the chatbot's features and services.
Sidebar Functionality
Search functionality within the sidebar to find specific chats quickly.
User Logout Option
Simple log-out feature for users to exit the chatbot interface securely.
Follow-up Questions
An AI chat offers follow-up question suggestions based on the user's initial inquiry.
Dashboard Widgets
Various widgets, such as Pie charts, Big Number widgets, Bar Char, Chat History, Interactive Dot Map, Interactive Treemap, and Manufacturers list, are available on the dashboard.
Comprehensive Information Provision
Commitment to delivering comprehensive information in response to user queries.
Chat Management
Users can create multiple chats. They can be efficiently managed from the sidebar.
Questioning Capabilities
Two request options: choose from suggested questions or enter a custom query (cost 1 credit).
Multilingual Support
UI elements are displayed in English. Multilanguage requests are responded to in the user's language.
Dashboard Generation
Automatic dashboard creation. When sufficient data is available, it is accessible via the "Show Dashboard" button.
Results
We have succesfully released an MVP version of the AI chatbot with the joint efforts of Orderfox, our team, and all interested parties. Now, it is available for beta users.
Our team is working on improving and scaling the chatbot. Further upgrades will be prepared based on the feedback of beta users, the team's investigations, market research, interviews with subject matter experts, and the application of practices and trends of the year.
The work with Rolique is productive. The quality is very positive and gives us flexibility on the topics we would like to improve or work on. We feel like you find solutions where others look for the problems only. We want to express our gratitude for the positive aspects of our collaboration and believe that addressing the areas for improvement will contribute to an even more successful partnership moving forward. Thank you for your continued dedication, and we look forward to achieving greater milestones together
Timur Göreci
Chief Revenue Officer
Let's engineer your breakthrough
79015, Smal-Stotskoho St. 1 Ukraine
50-062, Plac Solny 15 Poland