Description

Project Manager | HoopadVision | Nov. 2023 – Feb. 2025

       Core competencies: Project Management, System Integration, Process Improvements, Cross-functional Collaboration, Problem-Solving, Innovation, Jira.

·        Led the development and successfully delivered computer vision projects within the HBOX (the company’s product), including:

§ Smart Burglar Alarms: Improved system reliability and client satisfaction with a 70% reduction in false alarms by means of object detection.

§ Intelligent Parking: Accomplished License Plate Recognition (LPR) to work with mechanical gates to automate parking management for smooth entry and exit of vehicles.

§ Attendance Systems: A contactless attendance system based on face recognition for better efficiency at the workplace and security.

§ Smart Fire Alarm System: A proactive fire and smoke detection service that enhanced the safety of industrial and residential areas by 80% due to early fire and smoke alerts. (Link)

·        AIOT (AI + IoT): A smart home control system using Persian Voice Commands and a Chatbot powered with NLP and LLMs to provide smooth interaction for getting connected and controlling home devices. (Demo Link)

·        Developed and delivered all processes of the project, including RFP response, contract negotiation, and final delivery, which improved efficiency and ensured customer satisfaction.

·        Developed proposals, managed project schedules and cross-functional teams, mitigated risks, ensured effective client communication, and realized first-attempt project approvals with efficient FAT/SAT processes, comprehensive documentation, and rapid productivity ramp-up.

 

Computer Vision Engineer | HoopadVision | Jun. 2023 – Feb. 2025     

Core competencies: Software Development, Performance Optimization, System Integration, CCTV Config., Deep Learning, ONNX, Python OOP, R&D.

·        Improved Sales and Functionality in Surveillance Solutions: Enhanced the rate of product acceptance in the surveillance sector by 10%, identified key customer requirements, designed, and implemented all computer vision models and applications on Linux-Ubuntu to increase stability and security; used Docker containers for effective deployment and management:

§ 2x Faster Face Detection Service: Managed to reduce face detection inference time by 50% by rewriting the code and integrating ONNX, CUDA, TensorRT, and Nvidia DeepStream.

§ Biometric Liveness Detection: Developed a custom machine learning model for detecting spoofing attempts (an important concern within CCTV security systems) at an accuracy of 99%.

§ Face Super-Resolution: Developed a state-of-the-art deep learning model to enhance low-resolution CCTV faces; this leads to better recognition in challenging lighting conditions.

§ Vehicle Color Recognition: Developed a deep learning model to robustly recognize the colors of vehicles captured from CCTV.

·        Intelligent Parking System using License Plate Recognition (as an additional contribution): Designed a Python-based system for communication between a PLC (Modbus and Ethernet protocols) and an LPR system. This allows the gate to open automatically when an authorized car approaches and enables smart parking management. (Link)