Amir Esmaeili

Deployed back-end apps with German AI (TTS/STT) for real-time interaction, improving response time >3x via streaming. Refactored a high-load RAG web crawler using RabbitMQ & async programming, boosting throughput 4x and reducing latency. Developed an XR shopping assistant (Unity/Magic Leap II) with product recognition & indoor navigation for visually impaired users (400m²). Optimized CI/CD pipelines, doubling delivery speed and enhancing production stability.

* Migrated a legacy Flask application to Django with MySQL, resulting in a 30% improvement in system performance and enhanced scalability. * Leveraged Redis caching to store temporary data, reducing database queries 20% per prediction, and enhancing the performance of prediction operations. * Integrated RabbitMQ for real-time data processing, managing sensor data from 50+ distributed Raspberry Pis.

* Authored an advanced C++ and OpenCV plugin to stabilize Microsoft HoloLens II holograms, achieving a 100% reduction in mean and 60% in the standard deviation of movement-induced jitter. * Collaborated on Unity library integration, enabling real-time tracking at 15-20 FPS with under 6mm error.

* Contributed to the design and implementation of a Python/Django back-end with Angular front-end, processing 200 TB of satellite images to deliver real-time farm analysis under 150 ms. * Enhanced machine learning modules for crop detection and farm analysis, improving performance 4×. * Converted the DBSCAN project from Matlab to Python and enriched it for an 80% speed improvement.