○ Establishing functional requirements with our clients, using Google Suite and Notion as our main collaboration tools.○ Defining the system architecture based on the Clean Architecture principles.○ Creating a mock server which provides predefined request and response data to help our client's technical teams integrate our tool while development is ongoing.○ Developing a tool that, given a well-defined Vehicle Routing Problem dataset, compares the results of different VRP solutions, primarily our in-house solution Opticore, with other open-source solutions like Google OR-Tools, Vroom, Graphhopper Route Optimization API, andArcGIS VRP API.○ Implementing a scheduler that periodically runs trip optimizations. The technologies primarily used were TypeScript, MongoDB, PostgreSQL, MikroOrm, Redis, OSRM (Open Source Routing Machine).○ Conducting code reviews and making technical decisions.○ Developing several services for the project, such as the Driver Availability Service, Deliveries Service, Delivery Booking Service..○ Assisting in deployment using AWS CodePipeline, CodeBuild, and Beanstalk. The technologies used included Go, TypeScript, React, Puppeteer, MongoDB, Rollbar, and NewRelic.○ Developerd a Routing Problem solver in Go, functioning in two modes—light mode for fast response times (less than 500ms) used mainly to propose delivery time slots, and full mode for optimizing confirmed delivery distribution across different routes. It considers various constraints such as capacity, skills, time windows, costs, breaks, maximum travel time, and maximum load. My contributions included designing the architecture, setting up tests and test templates, implementing the first version of the light mode, and implementing the full mode using the open-source project VROOM. The technologies used were Go, Fiber as an HTTP framework, and MongoDB as a database for managing optimization jobs and their results.

● I Led a team consisting of two developers, a data scientist, a QA engineer, and a DevOps engineer working on Transport Management System. We utilized TypeScript and Express.js, adhering to the Clean Architecture principles. The project was deployed using Kubernetes and provisioned through AWS EKS.● I developed a driver application using React Native, which incorporated a variety of features including OTP authentication, real-time position tracking, delivery itinerary, and tracking interactions with end customers via SMS and calls, among others. The technologies I used for this project included ReactNative, TypeScript, MapboxGL, Mirage.js for mocking the API server, and Jest for testing.● Developed a geocoding microservice based on OpenStreetMap data, in Geojson format, and other open-source projects built on top of OSM, namely Nominatim , Photon, and Osmium. This microservice was implemented using TypeScript and MongoDB. Its main functionalities are: ○ Geocoding ○ Reverse Geocoding ○ Routing between locations by computing distances and durations ○ Management of Zones, which are polygons that represent clusters ofdifferent locations

● I created a suggestion system that informs drivers of the best locations to be around, at specific dates and times, implemented using k-means clustering, JS and MongoDB. These computations are performed periodically and are managed by the job scheduling library Agendajs.● I led and highly contributed to the transformation of our core product, which included a backend, two web applications, a Chrome extension, and two mobile apps, into a SaaS product. My contributions included: ○ Creating a role and permission-based authorization system. ○ Developing a wallet payment system that allows B2B clients to add credit points to their accounts and use these points for various functionalities. I utilized JavaScript, MongoDB, Mongoose, and the Tap online payment service. ○ Adding a page to the back-office application that displays all transactions made through the wallet system using React, Redux, and Websockets.● Created a forward/reverse geocoding microservice using TypeScript, MongoDB, and the Google Maps API. This microservice encompasses several functionalities: ○ Automated the extraction, cleaning, and formatting of Geographic Information data from PACI (Public Authority for Civil Information) in Kuwait. ○ Implemented forward geocoding in two ways: based on full-text search and address component- based methods. ○ Implemented reverse geocoding. ○ Managed failed geocoding attempts with a user-friendly interface.● Led the migration of an app used for placing personal delivery orders from React Native to a PWA (Progressive Web App).● Create an invoicing microservice that generates invoices for B2B clients, manages payments, notifications and reminders.● Developed a pricing engine using TypeScript. This engine calculates delivery prices based on complex criteria, including client type, delivery type, package type, delivery distance, and duration.

● Created a cross-platform mobile app using React Native that enables end customers to place personal delivery orders. I utilized Google Maps for mapping and FCM/APNs for notifications.● Developed an alerting system within a backstage web app that notifiesadmin officers when drivers are running late, deviating from their routes, ornot accepting delivery requests. This system was implemented using websockets via Pusher, with JavaScript, Go, and MongoDB.● Designed and implemented multiple charts to visualize key business KPIs within the backstage environment, using JavaScript, React, and recharts.● Created a dashboard for monitoring and managing all active orders and deliveries, utilizing JavaScript, React, Redux, and Google Maps. The dashboard was deployed on Heroku.● Maintained a legacy Android mobile app used by delivery drivers usingKotlin, FCM and Websockets.