Mabaat - End-to-End Booking & Property Management

Background & Problems

Mabaat is a comprehensive property and booking management platform designed to support accommodation reservations, payments, and operational workflows, primarily serving properties in Saudi Arabia. As the platform scaled and integrated with multiple third-party services, it encountered significant challenges related to payment reliability, data consistency, and post-payment booking management.

Integrations with external APIs introduced complex edge cases, such as incomplete data, inconsistent callbacks, ambiguous booking states, and limited visibility across the booking lifecycle. Furthermore, the platform struggled with proper RTL Arabic language support. This caused layout inconsistencies and localization bottlenecks, which directly affected operational efficiency, partner trust, and the user experience for the local demographic.

Objectives

• Stabilize booking and payment flows across all scenarios
• Ensure consistent and traceable data synchronization
• Support multi-channel bookings without overbooking risks
• Build a scalable, testable, and maintainable platform foundation
• Deliver flawless RTL support for the target market

The Process

The project began with a full audit of existing booking, payment, and integration flows. User journeys, API payloads, callbacks, and database states were mapped end-to-end to identify failure points and unclear transitions.

A flow-first approach was applied, defining explicit system states before, during, and after booking and payment, including error and retry scenarios.

Development Phase

• Refactored booking and payment flows with explicit state management
• Implemented defensive validation to prevent empty or partial data
• Added detailed logging and error tracing for faster debugging
• Standardized FE-BE data contracts for predictable integrations
• Prepared comprehensive API testing scenarios

Each enhancement was carefully evaluated against delivery timelines, recognizing that changes in payment or booking logic directly impact system stability.

Third-Party Booking Channels & OTA Integration

Mabaat integrates with multiple third-party booking channel APIs, allowing properties to receive reservations from external platforms while maintaining centralized control. Availability, pricing, and booking data are synchronized to minimize inconsistencies and overbooking.

A normalization layer was implemented to unify varied OTA payloads into a consistent internal format, enabling reliable downstream processing.

Property & Reservation Capabilities

• Check-in / Check-out Management: Configurable rules, time windows, and property-specific instructions
• Promotions & Discounts: Flexible campaigns based on dates, availability, and marketing needs
• Amenities Management: Structured amenity listings for transparency and better guest decisions
• Nearby Locations: Highlighting attractions, landmarks, and facilities around properties
• OTA Support: Unified handling of internal and external bookings

Location, SEO & Payments

Google Maps integration enhances location accuracy, navigation, and discovery by displaying properties alongside nearby points of interest. SEO best practices were applied across property pages using structured URLs, optimized metadata, and standardized property naming formats.

The payment system supports multiple methods, including Apple Pay, delivering a fast and secure checkout experience while maintaining robust post-payment booking management.

Delivery & Localization

Mabaat was delivered as a complete ecosystem to serve all stakeholders. To address the critical RTL Arabic language requirements, we implemented robust RTL support across all platforms. We integrated Tolgee, leveraging its AI capabilities to automate and manage translations seamlessly. This approach significantly reduced localization time and ensured contextual accuracy for Arabic terms, maintaining structural integrity across the following interfaces:

• Mobile Applications: For guests to browse, book, pay, and manage reservations
• Public Website: SEO-optimized for discovery and direct bookings
• Owner Dashboard: For property owners to manage listings, pricing, availability, promotions, and bookings

All platforms share a unified backend architecture, ensuring consistency, scalability, and reliable data flow.

Deployment

Deployment was executed incrementally using feature flags and controlled rollouts. Transactions were closely monitored in real-world conditions, with rollback strategies prepared to ensure business continuity.

Documentation

• System architecture diagrams
• Booking and payment flow diagrams
• Property naming and formatting standards
• Third-party API integration guidelines
• Error-handling and known edge cases

Impact

For Users in Saudi Arabia

• Highly intuitive and accessible experience tailored to the local community through seamless RTL support, accurate AI-assisted translations, and smooth navigation across all devices

For the Client

• More reliable booking and payment operations
• Improved partner and customer confidence
• Reduced operational issues and manual intervention

For Developers

• Cleaner and more maintainable architecture
• Faster debugging and issue resolution
• Lower risk of regressions as the platform scales

For Designers & Product Teams

• Clearer and more consistent user journeys
• Better UX decisions based on defined system states
• Stronger alignment between product vision and technical execution

Logie - Streamlining Global Campaigns

Background & Problems

Logie is an omnichannel influencer platform designed to streamline influencer marketing campaigns across multiple social and e-commerce channels. As brands increasingly leveraged creators for digital visibility, they struggled with fragmented workflows, slow campaign setup times, and inconsistent performance tracking across platforms.

The industry lacked an integrated solution that could unify influencer discovery, campaign execution, and analytics in one place, leading to inefficiencies, prolonged launch times, and data silos.

Objectives

• Create a unified platform for influencer discovery, campaign orchestration, and analytics
• Enable brands and creators to collaborate seamlessly across channels
• Build robust API integrations for reliable data flow between platforms
• Reduce time-to-launch for influencer campaigns

The Process

The development began with comprehensive user and market research to understand pain points experienced by brands, marketing teams, and influencers. This phase included interviews, competitive analysis, and product workshops with stakeholders.

A modular architecture approach was chosen so that individual platform capabilities, such as discovery, campaign management, and analytics, could be developed, tested, and scaled independently while maintaining consistent UX across the product.

Development Phase

• Designed a scalable microservices-based backend to support high throughput and extensibility
• Built frontend interfaces using React.js for dynamic and responsive interaction
• Integrated GraphQL for flexible, efficient data querying across client applications
• Implemented real-time notification and activity feeds for up-to-date campaign status
• Developed secure role-based access control for both brands and creators

Key Integrations

Logie integrates with major third-party platforms, including:

• Amazon API: For product-driven influencer campaigns and attribution
• TikTok API: For creator discovery and performance insights
• Social platform APIs for cross-channel publishing and analytics

These integrations allow automatic retrieval of content performance metrics and campaign ROI calculations in near real-time, eliminating manual data aggregation.

Feature Highlights

• Influencer Discovery: Search and filter creators by audience, engagement, and niche
• Campaign Builder: Intuitive UI to configure campaigns with automated channel rules
• Analytics & Reporting: Unified dashboards to compare engagement and conversions
• Collaboration Tools: Messaging, task assignments, and approvals within campaign workflows

Deployment

The platform was deployed using containerized infrastructure with CI/CD pipelines to support rapid releases, automated testing, and rollback strategies. Monitoring and alerting were implemented to maintain uptime and performance SLAs.

Documentation

• System architecture diagrams
• API contract specifications for internal/external use
• Frontend component libraries and UI guidelines
• Onboarding and developer setup guides

Impact

For Clients (Brands & Agencies)

• Reduced campaign setup time by over 50%
• Unified performance insights across all channels
• Improved ROI tracking and data-driven decisions

For Developers

• Modular, scalable architecture reduced maintenance overhead
• API-first design allowed repeated integration use and extensions
• Improved testing coverage and deployment confidence

For Product & Design Teams

• Consistent UX patterns led to higher user adoption
• Clear component standards improved design collaboration
• Modular UI allowed faster feature iterations

Grannfordon - Modernizing Car Sharing

Background & Problems

Grannfordon is a smart vehicle rental platform designed to modernize car sharing and short-term leasing across Scandinavia. Traditional vehicle rentals often rely on manual handovers, physical keys, fragmented verification processes, and limited real-time visibility into vehicle status, creating friction for both renters and vehicle owners.

The challenge was to build a fully digital, keyless experience that supports secure identity verification, real-time vehicle control, transparent rental flows, and scalable operations for multiple stakeholders.

Objectives

• Enable fully keyless vehicle access and control
• Support flexible rental and delivery options
• Provide real-time vehicle monitoring and tracking
• Build a unified ecosystem for renters, owners, and operators
• Ensure secure authentication and payment flows

The Solution

Grannfordon was delivered as a complete digital ecosystem consisting of a public website, mobile applications, and an owner dashboard. The platform enables renters to browse vehicles, estimate rental costs, book instantly, and access vehicles without physical keys.

Platform Components

• Mobile Apps: For renters to search, book, unlock, and control vehicles
• Website: For discovery, onboarding, and rental flow entry
• Owner Dashboard: For hosts to manage vehicles, pricing, availability, and rental history

Key Features

• Keyless Access: Engine on/off, lock/unlock via mobile app
• Vehicle Monitoring: Tire pressure, battery level, and vehicle status
• Flexible Pickup: Vehicle delivery or pickup at host location
• Rental Estimation: Transparent cost estimation before booking
• Live Tracking: Real-time location and trip visibility

Third-Party Integrations

The platform relies on multiple external integrations to ensure security, reliability, and automation:

• Smartcar API: Keyless vehicle control and telemetry data
• Google Maps API: Location tracking, routing, and distance estimation
• Stripe API: Secure payments and transaction handling
• Bolagsverket API: Vehicle and company data validation via plate number
• BankID SSO: Secure identity verification for Scandinavian countries

Development Phase

• Designed API-first backend to support mobile and web clients
• Implemented secure role-based access for renters and owners
• Handled real-time vehicle state synchronization
• Built fault-tolerant integrations with external APIs
• Ensured compliance with regional security and identity standards

Deployment & Operations

The system was deployed with scalability and reliability in mind, supporting concurrent rentals, live telemetry, and secure financial transactions. Monitoring and logging were implemented to track vehicle activity and user behavior.

Technology Stack

• Backend: Node.js
• Web Frontend: Next.js
• Mobile Apps: Flutter
• Database: PostgreSQL

Impact

For Renters

• Seamless keyless rental experience
• Full transparency on pricing and vehicle condition
• Convenient pickup or delivery options

For Vehicle Owners

• Centralized dashboard for vehicle management
• Reduced operational overhead
• Improved utilization and visibility

For the Development Team

• Clean separation between platform components
• Scalable architecture for future growth
• Strong foundation for additional mobility services

Bidubadu - Auction and Marketplace Platform

Overview

Bidubadu is a real-time auction marketplace built specifically for gaming collectibles, action figures, and rare enthusiast items. The platform connects collectors and buyers globally, enabling transparent, competitive, and engaging bidding experiences across web and mobile.

Our team oversaw the full development lifecycle, delivering a seamless experience for sellers, buyers, and administrators across Web, Android, and iOS platforms.

Objectives

• Create a dedicated auction marketplace tailored for collectibles and rare items
• Ensure real-time bidding with instant updates and accurate auction states
• Deliver a clean, intuitive UI that avoids confusion during high-pressure bidding moments
• Support secure payments, order processing, and shipping workflows
• Scale reliably for concurrent users and live auction events

Key Features

• Live Auction System - Real-time bidding, countdown timers, and instant winner determination
• Seller & Buyer Dashboard - Manage listings, bids, transactions, and order status
• Admin Dashboard - Full control over users, auctions, disputes, and marketplace activity
• Secure Payments - Integrated with Midtrans Payment Gateway
• Shipping & Order Tracking - From auction win to delivery confirmation
• Cross-Platform Access - Web, Android, and iOS applications

Technical Challenges

The biggest challenge in Bidubadu was ensuring that every critical marketplace interaction operates in real time. Auction systems demand accuracy and fairness, where delays or inconsistencies can severely impact user trust.

• Synchronizing live bids across multiple users and devices
• Preventing race conditions during last-second bidding
• Ensuring auction results are final, consistent, and tamper-proof
• Maintaining performance during peak bidding traffic

Development Approach

We designed Bidubadu with a clear separation between business logic, real-time events, and presentation layers. The system ensures that auction states, payment status, and shipping workflows remain consistent across platforms.

• Event-driven architecture for bidding and auction state changes
• Optimized database queries to handle high-frequency bid updates
• Carefully crafted UI flows to minimize cognitive load during auctions
• Consistent UX patterns across web and mobile apps

UI / UX Considerations

Auctions can be stressful for users, especially during final moments. We focused heavily on clarity, visual hierarchy, and feedback.

• Clear bid status and countdown indicators
• Immediate visual feedback after placing bids
• Simple, distraction-free auction pages
• Consistent interaction patterns for buyers and sellers

Platforms Delivered

• Web Application - Marketplace, auctions, dashboards
• Android App - Native mobile experience for bidding and selling
• iOS App - Optimized for iOS users with real-time interactions
• Admin Dashboard - Moderation, analytics, and system control

Technology Stack

• Laravel
• MySQL
• Midtrans Payment Gateway API
• Android & iOS (Mobile Apps)
• Web Application

Impact

Bidubadu successfully delivers a reliable, real-time auction experience tailored for collectors and enthusiasts. The platform enables secure transactions, competitive bidding, and scalable growth while maintaining clarity and trust throughout the user journey.

Psikologie - Psychological test Platform

Psikologie is a web-based psychological testing platform designed to deliver structured, reliable, and scalable online assessments. The platform enables institutions and organizations to conduct psychological evaluations efficiently while maintaining methodological accuracy.

Background & Problem

Traditional psychological assessments are often time-consuming, manually processed, and limited by physical presence. Educational institutions, clinics, and HR departments needed a digital solution that could deliver fast, accurate results without compromising established psychological methodologies.

Objectives

• Provide online and offline psychological assessments in education, clinical, and HR domains
• Deliver real-time test results using validated psychological formulas
• Enable HR departments to streamline recruitment and internal evaluation processes
• Ensure usability for both individual users and enterprise-level organizations

Key Features

• Real-time Test Results
  Automated scoring and evaluation using specific psychological calculation methods, allowing users to receive results immediately after completing assessments.
• Multi-Domain Assessments
  Supports psychological testing for education, clinical evaluations, and human resources.
• HR Subscription Model
  Companies can subscribe to HR plans, enabling HRD teams to conduct direct recruitment, staff screening, and internal psychological evaluations.
• Online & Offline Assessment Support
  Flexible assessment modes to accommodate various testing environments and operational needs.

System Design & Development

The platform was built with a focus on performance, data accuracy, and maintainability. Laravel was used to handle business logic, authentication, subscription management, and test workflows, while MySQL ensured reliable storage of test data and evaluation results.

Impact

• For Organizations: Faster recruitment cycles and standardized psychological evaluations
• For Professionals: Reduced manual processing and improved assessment accuracy
• For Users: Immediate feedback and a seamless testing experience

Technology Stack

• Backend: Laravel
• Database: MySQL
• Architecture: Web-based, modular, and scalable

VictoryToys - Action figure and Statue Marketplace

Background & Overview

VictoryToys is a specialized e-commerce destination catering to high-end collectors of action figures and statues. The primary challenge was creating a platform that handles high-traffic limited releases while maintaining a premium feel suitable for luxury collectibles.

Key Features Implemented

Results

The platform successfully launched with a 40% improvement in load times compared to the previous legacy system. The automated payment gateway reduced manual verification tasks by 90%, allowing the client to focus on sourcing new products.

SimobiPlus: Digital Transformation for Bank Sinarmas

Background

Simaspay is a comprehensive e-money application developed for Bank Sinarmas in 2018. It was designed to provide a modern digital wallet experience, allowing users to handle daily financial transactions entirely from their mobile devices.

Core Financial Services

• Check Balance: Real-time access to digital wallet funds.
• Transfer: Seamless money movement between users and bank accounts.
• Purchase: Buying digital products such as airtime, data, and tokens.
• Payment: Integrated bill payment system for utilities and services.

Innovation & Specialized Features

• Pay by QR (QRIS): Early implementation of QR-based payments, now standardized as QRIS, enabling touchless merchant transactions.
• Cash Withdrawal: Secure functionality allowing users to withdraw e-money funds at designated points or ATMs.

Technical Role

The project involved complex integration with banking ledgers and high-security payment gateways to ensure every transaction was encrypted, traceable, and compliant with financial regulations.

NodPay: Simplifying Social Transactions

KreditPro: Financial Inclusion for SMEs

Summary

Streamlined the loan application process to under 24 hours through automated data verification modules.

Al-Hadi: Bridging Technology and Spirituality

Development

Focused on high-performance Arabic text rendering and efficient SQLite indexing for instant search across thousands of verses.

PDAM OCR: Automating Utility Billing

Technical Detail

Implemented on-device Tesseract OCR to allow meter readers to capture data in areas without internet connectivity.