Mission Summary

1. Creation of a Ride Assignment Algorithm

• Objective: Solve the planning issues for pre-booked reservations for a fleet of medical taxis, considering round trips.
• Achievements: Designed an algorithm that accounts for real-time traffic, data cost optimization, relevance of distances (distance to the ride and the ride itself), and delay tolerances to manage return trip adjustments. The algorithm adapts to real-time changes to optimize route efficiency.
• Challenges:
  • Order of assignments to ensure optimal relevance
  • Costs of real-time traffic API
  • System robustness for large schedules

2. Implementation of Caching Systems for Traffic Management

• Objective: Optimize traffic data management to reduce response times and costs while maintaining accurate forecasts.
• Achievements: Created two levels of cache:
  • A long-term cache providing imprecise data for large-scale predictions.
  • A short-term cache with more precise data for real-time ride management.
• Challenges: Response times of caching systems

3. Creation of a CRM for Fleet Management

• Objective: Enable fleet managers to supervise taxis, make manual adjustments to ride assignments, and manage billing.
• Achievements: Developed a comprehensive CRM including a transport voucher management tool, an automated billing system, and manual correction tools to improve flexibility in route management.
• Challenges: Fine-tuned UX with client operators for better understanding of the application and its capabilities

4. Integration of an LLM Model for Data Entry Simplification

• Objective: Automate and simplify repetitive data entry for rides, patients/clients, and transport vouchers.
• Achievements: Implemented a language model (LLM) to provide automatic completions and reduce the time required for managing reservations and client records.
• Challenges: Prompt engineering

5. Cloud Infrastructure with AWS

• Objective: Design a scalable and high-performance infrastructure to host the application.
• Achievements: Deployed the infrastructure on AWS EC2 for hosting services, with MongoDB as the main database and DynamoDB for cache data management.
• Challenges: Need for robust infrastructure for a critical application

6. Automated Tests for the Assignment Algorithm

• Objective: Ensure the robustness and reliability of the algorithm in various scenarios.
• Achievements: Implemented automated tests to evaluate the algorithm's performance under different traffic and scheduling constraints and configurations.
• Challenges: Ensuring comprehensive test coverage and reliability