Faiz Muhammad Email & Phone Number

Helsinki Metropolitan Area

AILiveSim, a Helsinki-based startup, uses game technology to create realistic, dynamic simulations for training and testing AI algorithms. Our software can augment datasets, collect associated ground truth, and train new learning algorithms using a curriculum, as well as conduct various types of testing- Generated synthetic data resembling real datasets by tweaking parameters in 3D environments.- Developed an object detection model for the maritime environment using synthetic data.- Engineered a sophisticated decomposition technique within the ResNet model, focusing on the computation of projection matrices and orthogonal components tailored for PCA analysis.- Implemented camera calibration using single images from simulated data to predict horizontal field-of-view (H-FOV), Brown-Conrady distortion parameters, and calculate the intrinsic camera matrix (K-matrix).- Generated LiDAR point clouds and performed voxelization to enhance scene understanding in simulated environments.- Leveraged LLaVA/LLaMA large language and multimodal models to automate scene descriptions and improve AI interaction with simulated scenes.- Successfully dockerized machine learning inference models for optimal resource utilization and isolation.- Deployed models as AWS Lambda functions, enabling scalable, serverless execution with reduced operational overhead.

Tampere, Pirkanmaa, Finland

Teaching Assistant, Pattern Recognition and Machine Learning Course- Conducted regular weekly sessions focused on assisting students with assignments in key areassuch as neural networks, decision trees, Bayesian learning, and reinforcement learning.- Responsible for evaluating and grading student assignments, providing constructive feedback tosupport their learning and understanding of complex machine learning concepts.

Turku, Southwest Finland, Finland

Amplon is an AI-assisted strategic planning and execution software that helps you and your teams make strategic plans faster.- Used natural language processing (NLP) to enhance the business objectives defined by the clientsin Amplon’s Hoshin Kanri software.- Developed and deployed a REST API that suggests improvements in business objectives on GoogleCloud Run.

The Valley, Anguilla

Ladar is an innovative light-based sensor technology for the maritime and offshore industry. The Ladar system detects, characterizes, classifies and tracks various types of objects in real time, ranging from humans to ice floes, floating debris and oil spills.- Designing end-to-end machine learning pipeline for surface level and underwater marine object detection.- Developed synthetic RGBD data using 3D rendering in Blender.- Explored open-source libraries, such as BlenderProc for RGB and Depth image segmentation.- Developed a system to detect motion in live camera feed and save frames that have valuable information. - Developed convolutional neural network models in PyTorch to generate a LIDAR (Depth) image from an RGB image and a Near Infrared (NIR) Image from an RGB image.- Developed a 3D Mesh using RGB image and its corresponding depth image in Plotly Dash.- Modified YoloV5 model to train on 5 channels: RGB + Lidar (Depth) + Infrared (IR) (this feature fusion resulted in higher precision and recall on the validation set)- Developed a data visualization interface using Dash to display model results.- Set up the environment to collect visual and thermal data from IP cameras installed in Oslo, Norway.- Implemented a motion detection algorithm from scratch to detect ship movement in the visual and thermal camera recordings (this was done to save only relevant frames in the database).Note: You can use the trained models for image modification, RGB to NIR Conversion, RGB to LIDAR Conversion and 3D Mesh generation on your dataset here: http://ai.ladar.co.uk/

Lahore, Punjab, Pakistan

TechnoJawan is a social initiative intended to train the deserving youth of Pakistan on digital skills that can enable them to create earning opportunities online. - Collected and repaired a total of 14 laptops to ensure their functionality.- Designed a digital literacy learning module aimed at fostering technical skills in students.- Delivered the digital literacy modules to deserving students attending trust schools located in Kasur and Lahore.- Contributed to promoting digital literacy in underserved communities, empowering students to develop technical proficiency and pursue academic and professional opportunities.

Offshore Navigation Ltd. (ONL) offered me a full-time job as a project analyst after seeing my progress as an intern.- Working on different sub-projects that include optimization of VoyOpt which is a sail planning system developed by ONL.- Providing support in devising strategies in close liaison with the Machine Learning team to improve the accuracy of weather data

Offshore Navigation is based in London with clients all over the World. They provide maritime solutions, such as vessel route optimization and fuel efficiency. - Worked on geospatial data wrangling and analysis.- Worked on various Python libraries, such as Xarrays and NetCDF4. - Explored heat maps of vessel positions on Marine Traffic.

Lahore District, Punjab, Pakistan

- Created assignments and graded quizzes for students enrolled in BSCS 6th semester. - Conducted TA sessions to discuss and resolve course related ambiguities with students.

Lahore, Pakistan

AGRIASAN was part of Plan9's Cycle'13. It is a multi-platform application that uses deep learning to diagnose multiple diseases in wheat crops. - Worked on data pre-processing and hyper-parameter tuning of their models- Used transfer learning by incorporating architectures like VGG16, VGG19 and MobileNet to improve the accuracy of the models

Lahore, Pakistan

- Developed informative articles covering the latest advancements in technology.- Ensured all written materials were accurate, engaging, and accessible to a wide range of audiences.- Collaborated with subject matter experts to research and gather information for articles and blog posts.- Conducted thorough reviews and revisions of all content to ensure adherence to style, grammar, and formatting guidelines.

Lahore, Pakistan

Afiniti is an American multinational data and software company that focuses on developing artificial intelligence for use in customer call centers. - Developed an index parser for artificially intelligent paired call log files. - Worked with C# and MySQL to develop two desktop applications for Afiniti’s Software Integration Team.

Master'S Degree, Data Science, 4.77/5.00

Bachelor'S Degree, Computer Science, 3.52/4.00