Loay Amin

• AI practice lead utilizing GenAI capabilities to empower humans build faster, better and more efficient solutions across cutting edge industries. Responsible for technical team coordination and product roadmap creation. Utilizing GCP AI Infrastructure (Vertex AI) for product delivery. Responsible for solution architecture from ideation to realization and production. Responsible for client facing consultation, demo presentation and follow ups.
• Technically Leading a squad of 25 engineers from AI, Full-stack, Devops, Testing expertise for GenAI Product deliveries serving UK and Gulf end PwC customers.
• Guided and influenced partners on recommended upgrades and enhancements to integrated AI solutions.
• Worked with customers or prospective customers to develop integrated solutions and lead detailed architectural dialogues to facilitate delivery of comprehensive solution.

• Built Marvin; Udacity’s first GenAI bot for helping students with a personalized learning experience.
• Responsible for building automation tools for student and mentor operations using AWS, Google Sheets, docker, cloudformation, terrraform. Acting as cloud architect for a automated filtration and assessment process for mentor performance and student project submissions. Tool has successfully reduced cost of mentor and student operations by 8% across global operations.
• Responsible for content creation/tailoring and servicing for DECI program; One of the biggest programs for the company aiming to enhance Egypt's technical foundation. By building technological tracks such as Data Science, Web development, Cybersecurity, Embedded Systems and Digital Arts. It's the only and biggest initiative of it's kind in the world. Currently working on expansion in GCC.

Took part in the early team building process and in building a cloud-based proctoring system that is used by 2 of the world's biggest EdTech companies.

• Built cloud-native and containerized application that's able to host tens of thousands of real-time users at the same time and is built on AWS Kubernetes EKS using industry standard tools such as FastAPI, Docker, Terraform, Elasticsearch, Gstreamer, Poetry, Postgres, Kafka, DVC, Grafana, Snowflake.
• Platform uses Object detection, tracking, segmentation, gaze and pose estimation to track students taking an exam. Tailoring all these AI models to work in sequence for thousands of students at the same time was successfully done through industry standard approaches.
• Develop a platform for monitoring, logging, caching, data layer, and orchestration of Docker containers running on Kubernetes.
• Worked on building CI/CD pipelines, unit-tests, integration tests and QA. All automated and seamless across team members.
• Created a dedicated CLI tool that's main purpose of to streamline the dev process between team members and automated all repeated sub tasks such as ticket/branch/merge creations, rebasing, Docker image version synchronization, infrastructure-as-code templates to help build new components..etc.
• Gained strong leadership skills by managing projects from start to finish.

Responsible for full deployment of several projects that involve Lidar technology, supported by Camera input and other auxiliary inputs. Our patented systems won several tenders with different Police departments in UAE and Egypt .

• P3-Speed: Moving speed enforcement system developed for Abu Dhabi Police. Produced for more than 220 cars at phase 1. This US patented software accurately detects moving vehicles on road (while moving), detects all speed/seatbelt and distracted driver violations and captures images of vehicle plates all in realtime. Currently, P3 is deployed in Egypt.
• I-Trac: Security system for VIP/blacklist people tracking, counting, analytics, PTZ tracking and other customized features for highest profile locations in Dubai such as CityWalk and DXB airport terminal 3 and currently Dubai Consulates and Graveyard area.
• Slopes: Road surveillance system that works on accurately measuring side road slope angles for road construction work. This project is developed for Ministry of Transportation in Qatar and has significantly reduced cost of rail road construction for Qatar highways.
• TRC: Total Road Control unit developed for Dubai Police in Alkheil road, for accurate speed enforcement, distracted driver and seatbelt violation safety for 8 lanes in both direction (total 16 lanes).

• The team's main scope is leading the AI vision of the one of the biggest car manufacturers in the world, Valeo. Took part, as a starting member, in researching state of the art techniques regarding autonomous driving, hardware deployment of complex models on limited hardware (cars) while maintaining real-time performance.
• Deep Learning Researcher at Valeo. Works on data extraction, analysis and post processing for safety-critical automotive projects (Regression, Semantic segmentation and detection). Performed data augmentation for better training accuracy.
• Responsible for reaching real-time performance for end-to-end autonomous driving car systems. Handles deployment for new state-of-the-art network architectures, profiling and benchmarking on ECUs like Jetson Board (Nvidia) and Movidius (Intel).
• Working on code optimization for kernel coding on the targeted ECUs using CUDA and openCL .