PparKing, a Swiss parking operator, relies on automated systems to detect vehicle occupancy and calculate parking duration. The platform enables parking operators to manage and bill parking sessions with minimal manual effort.

Customers can use the service through two payment paths:

• Advance booking via the company website using Twint, Switzerland's mobile payment solution

• Spontaneous parking without pre-booking: CCTV cameras capture license plates and dwell time, and the system generates invoices automatically

At the start of our collaboration, PparKing was already operating an existing Laravel application that processed CCTV images via FTP storage. It analyzed images to identify matching license-plate pairs, correlate entry and exit events, calculate parking duration, and generate invoices.

The legacy system suffered from significant performance bottlenecks. Its architecture relied on direct file-system operations for image processing, an inherently inefficient approach that caused major delays.

During the retainer period, we identified critical limitations in Laravel Nova (the original admin-panel framework). These constraints led us to migrate to FilamentPHP, our preferred admin-panel builder. We then rebuilt the application on top of the existing database infrastructure, which delivered three major improvements:

• Higher operational flexibility

• Measurably faster processing speeds

• Tailored views and interfaces for daily operations

This architectural switch from Laravel Nova to FilamentPHP enabled direct customer portal access and expanded self-service features for end users.

CCTV cameras at parking-lot entry points capture images when vehicles enter. Each entry image is registered as an event with timestamp, plate number, and direction metadata.

The same process runs at exit points. The system captures timestamp and plate number, matches image pairs, links them to billing records, computes parking duration from timestamp differences, and generates invoices based on the active tariff model.

Several automations run in this process. In Switzerland, registered vehicle owner data can be retrieved via municipal APIs. We built an integration that automatically queries municipal databases when retrospective billing is required.

With this platform, physical barrier gates are not required. Clearly visible signs at entry points define contractual usage terms. Camera monitoring starts as soon as vehicles enter the premises:

• Vehicles are photographed,

• license plate data is captured, and

• billing authority is established.

PparKing used industry-standard CCTV cameras with built-in OCR, operating as a closed hardware ecosystem.

We redesigned the workflow: instead of manipulating individual image files, the system now retrieves directory file lists and parses file names to extract plate data, timestamps, and metadata. Working with database records improved reliability while delivering major performance gains.

Once file metadata is catalogued in the local database, the system no longer depends on ongoing FTP storage reads. Whenever a new file appears, file-system notifications trigger automated checks:

1. Does a matching image exist for this file?

2. Is there a matching license plate?

3. Does it belong to an active parking session?

We built a custom validation interface that consolidates all session data in one view: entry/exit images, booking details, and parking metadata including site-specific grace periods. Color-coded indicators, green for paid and red for overstay, enable rapid case evaluation.

When OCR misreads occur, the system surfaces similar plate images so operators can connect fragmented data points and apply correct matches.

The primary challenge was the legacy system's weak performance and high error rate. The CCTV processing pipeline suffered from severe latency and reliability issues, requiring a full architectural redesign.

Handling missed plate detections and OCR misreads was especially demanding. Human oversight remains essential for edge cases that cannot be solved programmatically.

Secondary challenges centered on usability: simplifying workflows, reducing cognitive load, and removing unnecessary UI elements.

Customer complaint management system

We built an integrated ticketing system that displays email exchanges as conversation-like chat threads. Each invoice includes a system-generated authentication code. Customers access their case history through a secure portal using license plate and passcode, where they can review photo evidence, zoom images, and billing discrepancies.

Asynchronous messaging enables customers to submit requests at any time, while operator responses are sent during business hours.

Laramate GmbH treats comprehensive codebase assessment as a non-negotiable requirement before any engagement.

Our codebase audit is paid, non-refundable, and mandatory. We review security vulnerabilities, evaluate architectural integrity, identify technical debt, and audit third-party dependencies.

Would you buy commercial real estate without a building inspection? Yet companies routinely invest large budgets in software without understanding its technical foundation. Building on fragile foundations is a business risk.

Without an assessment, we cannot provide accurate scope, realistic cost estimates, or credible timelines.

Clients often contact us specifically because we already have answers to their core challenges.

We know how to structure Laravel applications for complex projects, implement bidirectional integrations, maintain long-term maintainability, optimize performance, apply proper security controls, and integrate Laravel with modern as well as legacy systems.