Laurentiu is a driven Machine Learning Engineer and Data Scientist with 6 years of expertise in automating processes for companies using tools like Python, Bash, Tensorflow, NumPy, Pandas, Scikit-Learn, Flask, and PyTorch. His work involves dealing with real-life business problems that require Deep Learning and Machine Learning techniques.
Use a combined knowledge of computer science and applications, modelling, statistics, and analytics to solve problems.
Administer infrastructure solutions to improve data models, increase data accessibility and foster data-driven solutions for clients.
Implement monitoring solutions to ensure data integrity - working closely with engineers, product managers and other stakeholders.
Performed data analysis using Pandas and built state-of-the-art models with PyTorch, TensorFlow, Sci-kit learn. Automated the surveillance system in bank branches by alerting people in distress, alerting crowded zones and queues, alerting unusual ATM session times, alerting CCTV malfunctions.
Created ML architectures and implementation plans. Trained and deployed on edge devices and cloud environments (Flask, Django, Azure, AWS, Docker, Kubernetes).
Designed and developed the pipelines (data acquisition, model inference, training, communication, business features) for Lummetry's main product, which is a distributed AI-based near-real-time safety & security system (computer vision).
SolisBox is an Enterprise AI solution for safety and security, integrated into an IoT platform.
It gathers data from CCTV cameras, video files, and sensors and analyzes it in near-real-time using deep neural models.
The solution employs Python, OpenCV, TensorFlow, PyTorch, and Numpy, along with Docker, GitHub, and Bash for CI/CD and communication.
A machine learning model was developed to identify duplicate payments.
Siamese networks were utilized to reduce manual labor by 90% while maintaining high recall.
The project's primary data scientist was responsible for analyzing data and providing a neural model solution.
Developed an 'as-a-service' solution for predictive analytics, integrating with customer data via API.
Responsible for data science and machine learning engineering for the project, including proof of concept.
The solution is now part of the SolisBox ecosystem, previously referred to as Project 1.
