Glinteco

Challenge

• Accurate, quick processing of up to 5 faces via computer cameras
• Requires the system to run smoothly right on the browser without server-side AI
• JavaScript is not optimized for heavy AI computing tasks such as image recognition
• Need to ensure high accuracy when recognizing across different races, ages and lighting conditions
• Resource limitations on devices such as digital signs or mini-PCs make model optimization a key factor

Solution

Glinteco developed a modular, high-performance browser AI app, focusing on user experience and scalability:

• Using TensorFlow.js to run face recognition models and predict age, gender directly on the browser
• Trains custom AI models on diverse datasets for high accuracy/reliability
• Optimizes models for lightness, high performance, reduced frame analysis latency
• Leverage browser hardware acceleration and parallel processing techniques to maintain high processing speed
• Apply performance programming strategies such as memory reuse, batch processing, and computational reduction techniques
• Build a modular architecture that allows easy model updates or integration of new features such as people counting, emotion detection, and behavior recording
• Simple interface, no backend, quick deployment to any browser-enabled camera system

Result

• Achieves 91% age/gender detection accuracy, even with 5 faces simultaneously
• Reduce processing time per frame by 60%, meeting real-time analysis requirements
• Enables deployment quickly into existing systems without backend infrastructure, reducing deployment time by 70%
• The system consumes very little machine resources, allowing continuous operation 24/7 without the need for intensive technical maintenance

Challenge

• SMEs in Singapore often face the following obstacles:
  Complicated accounting and financial reporting processes that require a lot of time and manual labor.
• Lack of tools to customize reports according to the specific business model.
• No overall system to integrate payroll, taxes, expenses and reporting into a single platform.
• Lack of data analysis capabilities leading to lack of information for business decisions and planning.

Solution

Glinteco and Elmlab develop Elmlab.my EOS with a dynamic and convenient digital platform approach:

• Building backend with Django, providing stable API for frontend deployment and data management process
• Designing easy-to-use UI/UX for both accountants and small business managers
• Integrating cost tracking, payroll management, tax calculation and automatic financial report generation
• Allowing users to create and customize report forms suitable for each business
• Background processing mechanism using Celery/Redis helps improve performance when businesses expand data
• Using PostgreSQL and cloud storage to store important data, ensuring safety and scalability
• Consulting on digital financial processes, helping business owners improve management efficiency and cost control

Result

• Reducing 70% of the time to prepare financial reports and manage payroll compared to manual methods
• Optimizing 60% of expenses and taxes through automatic alerts and recommendations
• Supported over 20 businesses using the platform in the first 6 months of launch
• All forms are customizable and can be printed/exported quickly, increasing accounting office productivity

Challenge

• Outdated interface, suboptimal user experience
• Slow loading page, poor performance
• Old backend system, difficult to maintain
• Unoptimized database and infrastructure
• Images do not meet modern aesthetic standards

Solution

• Redesign modern, easy-to-use UI/UX interface
• Clean up the entire source code, remove redundant code, optimize existing functions
• Move database and server infrastructure to a new, more stable and secure environment
• Optimize images according to web standards, ensure synchronization between aesthetics and page loading performance
• Check and patch all potential vulnerabilities related to backend security

Result

• Increase user satisfaction by 50% with the new interface
• Reduce page loading time by 60%
• Increase concurrent access handling by 100%
• Reduce backend errors by 70%, improve overall stability

Challenge

• Combine creativity in design with complex technical requirements
• Ensure fast loading speed and smooth experience despite integrating many animations, multimedia content
• Optimize security when handling user data and interactions on the web platform
• Balance between modern UI/UX and backend performance

Solution

• Use Django CMS to ensure flexible content management, easy updating, and high security
• Design user interfaces using HTML/CSS combined with lightweight JavaScript libraries, optimizing display without reducing loading speed
• Apply optimization techniques such as image size reduction, caching, and minifying source code
• Implement standard security measures such as CSRF, XSS protection, and access authentication
• Build a modern, smooth interface with light animations, consistent with the brand

Result

• Increase user interaction time by 65% thanks to engaging design lead
• 40% reduction in bounce rate
• 70% increase in page load speed compared to the first test

Challenge
Before Glinteco joined, the museum faced many challenges when converting from traditional exhibition format to digital platform:

Combining art and technology: How to express cultural and historical values in digital environment while still being attractive and easy to understand for all audiences.

• Friendly design, covering many audiences: Children, adults, tourists, students - each user group needs a suitable approach.
• Optimizing diverse content: Hundreds of artifacts, stories, documentary photos, videos, audio - all need to be presented scientifically and attractively.
• Loading speed and scalability: The website needs to load quickly even with hundreds of photos, interactive audio/video.
• Requirement for easy updating system: Museum team needs to proactively update new events and activities without technical knowledge.

Solution

• Glinteco has come up with a comprehensive solution to modernize the entire museum web experience:
• Unique UI/UX design: Combining heritage colors, light animations, interactive interfaces suitable for all ages, increasing attractiveness and ease of navigation.
• Building a digital exhibition system: Photo gallery, Exhibition with audio commentary, Interactive documentary videos, Virtual tour simulates the experience of going to a real museum
• Performance optimization: Optimize images, reduce JavaScript/CSS load, use cache and CDN to help the website load quickly and stably even with low connections.
• Easy-to-use content management system: Museum staff can easily update articles, news, events without knowing programming.
• Multi-device compatibility: The interface is compatible with computers, phones, and tablets.

Result

• 130% increase in traffic after 3 months of implementation
• 45% reduction in bounce rate thanks to intuitive interface and interactive content
• 60% increase in average time spent on site
• 70% reduction in time required for museum team to update new content

Challenge
Before implementing the project, Viet Plastic encountered many barriers in operation and development:

• Manual management: Most of the production, warehousing, product and order management processes were done through books or separate tools (Excel), leading to errors and lack of transparency.
• Lack of integration: Data is not synchronized between departments, making decision making slow and inaccurate.
• Limited access to the market: There is no effective portal to introduce products, making it difficult for businesses to expand their customer base and market.
• Not ready to expand: The current system cannot be upgraded to meet ERP needs or handle large production data (500+ tons of plastic/month).

Solution
Glinteco has built a comprehensive digital platform, from customer interface to internal administration:

• Django & Django CMS platform: Build a powerful, flexible, secure and scalable backend.
• ORM integration: Optimize data queries and transactions with easy scalability when the amount of data increases.
• VueJS administration interface: Increase interactivity, easy to use for content management, orders, products, warehouse operations.
• Beautiful frontend, good response: Use modern HTML/CSS/JS to create a friendly, SEO-standard interface, can display hundreds of products smoothly.
• Big data processing with Celery & Redis: Ensure the system runs stably when handling heavy tasks such as production statistics, real-time inventory updates.
• AWS S3: Store product images and documents effectively, ensure data security and reduce server load.

Result

• 70% increase in warehouse management and production process efficiency compared to manual methods
• 60% reduction in processing time for order and product related data
• 120% increase in website visits from enterprise customers
• System is stable when processing 500+ tons of production data per month

Challenge

Before implementing DKG Hospitality, DKG Group faced many difficulties in synchronizing customer experience and management processes:

• The booking interface was not optimized, causing disruptions in the service booking journey
• The old management system lacked flexibility, making it difficult to adapt when the number of hotels and guests increased
• Lack of customer behavior data analysis tools to make accurate business decisions
• Hotel staffs had to manually update room status, which was prone to errors and time-consuming
• There was no system to integrate marketing content or communicate effectively with potential customers

Solution

Glinteco cooperated with Dang Khoa Group to build the DKG Hospitality platform from scratch, focusing on user experience, content management and system performance:

• Designed an intuitive, modern user interface that matched the high-end brand image
• Allowed customers to browse rooms, view amenities and book easily from any device
• Developed the system Django backend combined with Django CMS, allowing real-time management of rooms, rates, statuses and transactions
• Using Django CMS to make it easy for management teams to create new content such as hotel introductions, services, images and promotions
• Integrating a blog module for hotel staff to share travel tips, promotions, and new updates - increasing customer interaction and supporting SEO
• Building a system to update room status in real time, reducing duplicate booking errors
• Authorizing users by role: group administrator, hotel staff, content editor
• Integrating a mechanism to collect user behavior data to support booking trend analysis, content effectiveness evaluation and personalization strategy

Result

• Increased online bookings by 75% within 3 months of implementation
• Reduced time spent updating room status and processing reservations by 60% for hotel staff
• Increased 80% of views on blog content and promotions pages on the system 
• Increase 45% of repeat bookings

Challenge

Challenges in building a large-scale HR management system, ensuring scalability and tight integration with other components in the GapoWork ecosystem

• Managing complex HR processes: employee records, timekeeping, payroll, KPI evaluation, leave, and onboarding/offboarding
• System stability for thousands of concurrent users, with fast response times and no data congestion
• Difficulty in synchronous integration with other systems like GapoWork Core, notification systems, task management
• Requirement for flexible, easily scalable system design, and intuitive interface meeting modern office user expectations
• Inability to use existing HR solutions due to limitations in customization and integration

Solution

Glinteco implemented a solution from scratch: modern architecture, high performance, comprehensive integration with the GapoWork ecosystem

• Golang & GIN backend: high performance, easy to maintain, handles millions of requests/day
• ReactJS frontend: flexible, smooth, easy-to-use interface with high interactivity
• Microservices: scalable for individual functions (payroll, KPI, employee management)
• OAuth2 integration: synchronizes user identity with the GapoWork ecosystem
• Core features: HR record management, contract information, compensation; Payroll system, tax deductions, and automated payslips; Customizable KPI performance evaluation; Real-time HR reporting; Work history, leave, and internal notification synchronization
• Flexible configuration mechanism: allows each company in the GapoWork ecosystem to adjust to its internal processes
• Continuous improvement and expansion of functions based on actual feedback from partner companies' HR teams

Result

• Increased HR administrative work efficiency by 70% compared to old processes
• Reduced monthly payroll and performance evaluation processing time by 50%
• Increased HR user satisfaction by 40% (internal survey)
• System stable for 10,000+ concurrent users
• Reduced manual processing needs by 60% thanks to automation 

Challenge
Developing a booking platform for a 1000+ room system requires complex data management and high performance.

• Handle large data/user interactions without compromising response speed
• Room status updates are not instantaneous, resulting in a less-than-smooth booking experience during peak hours
• An under-optimized frontend
• The system needs to support easy scalability

Solution
Glinteco has developed Paradisehome.vn with a solid technology architecture, combining modern tools

• Django backend for large transactions, fast data retrieval
• Integrating Django-CMS to manage room content, services and news, helping the operations team update information flexibly without programming intervention
• Celery/Redis for async tasks (room updates, emails), ensuring quick, uninterrupted system response
• VueJS frontend for smooth, real-time room/service status display
• Standard HTML/CSS/JS for responsive design across all devices
• Designing and integrating RESTful APIs for frontend and backend to communicate effectively, ensuring smooth data flow
• AWS S3/RDS for secure, highly available, scalable data/image storage
• Django ORM for fast, complex room/service data management

Result

• Increasing monthly online bookings by 25%
• Reducing user search and booking time by 60% thanks to optimized interface and response speed
• Increasing customer satisfaction by 40% through post-booking surveys
• The system is capable of displaying and effectively managing more than 1000 rooms, serving hundreds of simultaneous visits
• The reporting system is fast, timely, and real-time, helping managers grasp information and provide appropriate incentive policies.

Challenge

Endeverus.com faced key challenges in build

• Precise matching of business needs with student capabilities (project requirements, career aspirations)
• The system must be scalable to handle multiple projects, profiles and interactions simultaneously
• Manage asynchronous tasks such as sending emails, processing data, synchronizing documents effectively
• Establish stable real-time communication, serving instant chat and updates
• Store and display complex dynamic content such as project descriptions, documents, reports, videos, photos
• Ensuring an intuitive, easy-to-use interface for two distinct user groups

Solution

Glinteco has implemented a comprehensive solution, focusing on solid architecture, smooth user experience and scalable integration capabilities

• Django/Django REST Framework backend for stable transactions and large data processing
• Using Celery and Redis to handle background tasks such as sending emails, synchronizing, processing project data without interrupting the user interaction flow
• WebSocket is integrated to create real-time connections for features such as chat, update notifications, quick feedback between students and businesses
• The frontend initially used JavaScript and jQuery to quickly build interactive interfaces Upgraded to Next.js for improved performance, code reusability, and cross-device UX
• Markdown input converted to HTML/CSS/JS for dynamic content display (project descriptions, reports)
• Multimedia resources and user data are securely stored on AWS S3
• The interface is optimized according to the single-page application (SPA) design, reducing loading time, increasing response speed and inc

Results

• 35% increase in successful student-business connections every 2 months
• 50% reduction in average time for businesses to find suitable personnel
• Processes 3,000+ asynchronous tasks daily
• 40% savings on data storage costs

Challenge

DSTax.com was built to address major challenges in the field of tax integration and management:

• Lack of a centralized portal to standardize the tax calculation process
• Complex tax API integrations lacking unified structure
• Traditional tax declaration processes are manual, time-consuming, and error-prone
• The existing user interface is not user-friendly, making it difficult to operate
• Lack of flexibility and customization to expand integration or support many types of businesses

Solution

Glinteco developed a comprehensive, scalable, and user-centric platform to streamline workflows for tax professionals and integration developers.

• Build the backend with Django and Django REST Framework, creating a solid API foundation for deep integrations. Automatically store access history and tax requests.
• Use Celery and Redis to handle asynchronous background tasks, optimizing response speed and overall performance
• Choose PostgreSQL as the database management system to ensure integrity, stability, and easy expansion
• Successfully integrate with leading tax systems such as Sovos, ONESOURCE, and Vertex, making it easy to access, configure and send requests to tax service providers
• Develop API connectivity (OAuth2) with popular financial services such as Stripe, QuickBooks, ChargeBee, Shoplazza, ConnectWise, Xero, and Acumatica, for simplified customer integration
• The user interface is developed with HTML, CSS, JavaScript, and jQuery, using Ajax to improve the smooth, responsive experience
• Responsive design helps the system work well on all devices (desktop, tablet, mobile)
• Optimize UX with smooth transition and user-friendly layouts
• JQuery and Ajax code snippets help to quickly switch pages without reloading the entire page, creating a seamless feeling for end users

Result

• 80% increase in API integration efficiency (vs. manual)
• 60% reduction in average tax calculation time per session
• Achieved 99.99% operational uptime
• Capable of handling 1,500+ API requests per hour

Challenge

STLPro.com struggled with inefficient product data management and operational bottlenecks

• Product data was not updated in a timely manner (manually), affecting competitive pricing
• The rapid increase in product quantity and orders caused a significant decrease in system performance, delayed orders, and inaccurate inventory
• Data collection and synchronization are mainly done manually or semi-automatically
• Lack of tools for managing orders, customers, and shipping status
• No statistical system, profit analysis by product
• The current system cannot be expanded to other e-commerce platforms

Solution

Glinteco designed and implemented a comprehensive CRM + ERP system to automate operations, improve data accuracy, and optimize profit:

• Build internal CRM + ERP to manage customers, orders, shipping status, and transaction history
• Develop a buy bot to automatically perform actions such as ordering, payment, and tracking order status on suppliers' platforms
• Integrate Amazon Selling Partner API to automatically update order information, selling price, inventory, and product status
• Use Playwright, Selenium, and Puppeteer to collect data and perform automated browser tasks from other e-commerce sites
• Build a financial bot to automatically collect daily product data, analyze costs, revenue, and profits of each product, and automatically adjust selling prices
• Use Celery and Redis to handle thousands of background tasks every day with high performance
• Design a simple, clear admin interface that makes it easy for users to operate and monitor the system.
• Set up a monitoring and alerting system to detect errors early, automatically fix them, or notify administrators.

Result

• 85%+ operational process automated
• Product data processing speed increased 10x
• Manual operations and data entry errors reduced by 70%
• Order processing time and customer feedback reduced by 80%
• Financial bot increased average profit per product by 20 - 25%