Intro

Oh hello there! 😁

I am an Artificial Intelligence Specialist with several years of experience in Python, but for my friends, I’m still the person who speaks more in code than words πŸ˜‚

✨ My Passion and Purpose?

I’m passionate about AI because of its limitless potential to solve real-world problems and create future-changing technologies. I enjoy pushing boundaries with AI models, and I’m driven by the desire to innovate and simplify complex tasks through automation.

πŸš€ What I’m Proud to Have Accomplished:

  • AI Model Integration into Unreal Engine β†’ Obsidian Gateway – Developing a sophisticated pipeline for real-time AI vision integration into Unreal Engine, blending cutting-edge technology with immersive experiences.
  • Synthetic Medical Image Generation – Generating synthetic medical imagery that revolutionizes research and AI-driven solutions.
  • Abbreviation Disambiguation in Medical NLP – Enhancing medical systems' ability to interpret clinical abbreviations with precision.

πŸ’» I speak not in my mother tongue, but in Python! I’ve worked with a wide range of Python libraries and tools, mainly in AI, Networking, and Backend development. Thanks to my Master's in AI at Bologna, I’ve also gotten my hands on Prolog, Minizinc, and Lisp, enriching my toolbox for solving complex problems.

✏️ Alongside my work, I post educational content on LinkedIn and love sharing knowledge to help others grow.

πŸ‘₯ Let’s Connect:

I thrive on learning new things and am always eager to dive into fresh challenges. I’m very precise (some might say too precise πŸ˜‚) and I take deadlines seriously. I also have a habit of automating anything repetitive, making tasks a one-click affair.

πŸ‘‹πŸ» Thanks for visiting my website! Feel free to drop me a message anytime πŸ“§

Education

University of Bologna

Master of Science (M.S.) in Artificial Intelligence

  • Thesis title: Integration of AI models into Unreal Engine
  • GPA: 3.78

Sep 2021 - Oct 2024

University of Damghan

Bachelor of Science (B.S.) in Computer Science

Feb 2015 - Feb 2019

Work Experience

AWENTIA

Computer Vision Integration in Industrial Settings

  • Automated the complete AI pipeline for non-expert users across various industries, enabling businesses to adapt to the ever-changing nature of data efficiently.
  • Integrated AI models into edge devices, optimizing performance and reducing dependency on cloud-based solutions.

Nov 2024 - Present

NOVA XR

Integration of AI Models into Unreal Engine - Thesis

  • Engineered a low-latency pipeline that streams external camera feeds to a server for AI processing, achieving 20 ms model inference time and reducing network latency to 13 ms using WebSockets; enabled real-time integration of 30 fps input and output within Unreal Engine.
  • Utilized a range of tools and technologies, including C++, Python, JavaScript, along with Channels, Docker, PostgreSQL, Redis, and Nginx.
  • Completed several automation projects that reduced manual work by 90%.

Apr 2024 - Jul 2024

Teaching Assistant

  • Mentored over 300 AI students, improving their proficiency in Python, algorithms, and AI languages, leading to both academic and practical successes.
  • Collaborated closely with Professors Simone Martini and Michael Lodi, assisting in course material preparation and grading assignments, ensuring the efficient operation of the course.

Sep 2022 - Present

Freelance Python Developer

  • Integrated 50+ third-party APIs into Python applications, enhancing functionality and user experience.
  • Developed and deployed 10+ RESTful APIs using Flask and Django, improving data processing efficiency.
  • Implemented unit tests, increasing code reliability and reducing bugs.
  • Automated repetitive tasks, such as scheduling meetings from image-based event details, resulting in a 50% time savings in administrative workflows.
  • Optimized Python scripts, reducing execution time by up to 40%.

Feb 2019 - Aug 2021

Projects

Medical Abbreviation Disambiguation - NLP

  • Developed a medical abbreviation disambiguation system using negative sampling.
  • Fine-tuned TinyBert, BioBert, and SciBert models, achieving an F1 score of 0.81 with SciBert.
  • Processed and filtered a 5M-record MeDAL dataset, focusing on abbreviations with 20-30 expansions.
  • Performed error analysis and post-processing, boosting F1 score from 77% to 89% for key abbreviations.

GitHub Repo

Generation of Clinical Skin Images - CV

  • Developed and implemented a machine learning pipeline using ControlNet to train models on custom skin tone datasets.
  • Automated mask creation, skin tone extraction, and prompt generation for over 12,000 images.
  • Trained ControlNet on skin tone-based prompts, achieving Avg PSNR of 27.19, Avg SSIM of 0.67, and FID of 69.38, demonstrating high-quality image reconstruction.

GitHub Repo

Smart Form Auto-Filler Chrome Extension

  • Streamlines form filling by managing configuration data and user-specific prompts.
  • Integrates with OpenAI API to generate responses for incomplete fields.
  • Provides an easy-to-use form completion experience based on stored data.
  • Allows customization to meet specific user needs.
  • Uses Chrome Storage to keep data persistent across sessions.
  • Supports multiple types of form fields, including text, email, and dropdowns.

GitHub Repo

WebSocketActor - C++

  • Facilitates real-time WebSocket communication for Unreal Engine applications.
  • Handles both text and binary data, converting binary data into textures dynamically.
  • Applies textures to game materials in real-time for dynamic content updates.
  • Features on-screen debugging and logging for seamless development.
  • Easy integration into any Unreal Engine project with customizable connection settings.

GitHub Repo

Skills

Programming Languages

Programming Languages

Data Science & Machine Learning

Data Science & Machine Learning

PyTorch TensorFlow Pandas NumPy SciPy Keras Plotly

DevOps and Cloud Computing

DevOps and Cloud Computing

Development Tools & Environments

Development Tools & Environments

Version Control & Collaboration

Version Control & Collaboration

Databases & Data Management

Databases & Data Management

Operating Systems & Platforms

Windows 11 macOS Linux

Text Processing & Formatting

Text Processing & Formatting

Web Development & APIs

Web Development

Nginx Django Daphne Channels Django REST

Game Development

Game Development

Research Interest

My research interests span across a diverse range of topics in Artificial Intelligence, software engineering, and automation. I am particularly passionate about:

  • Computer Vision and 3D Reconstruction: Developing methods for image recognition, object detection, and 3D scene reconstruction to create immersive experiences in both medical and entertainment contexts.
  • Generative AI and Transformers: Leveraging Generative AI models, such as ControlNet and large language models (LLMs), to push the boundaries of creative content generation and enhance natural language understanding.
  • Natural Language Processing (NLP) in Medical Context: Building NLP systems to handle medical terminologies, improve communication in healthcare settings, and tackle the challenges of disambiguation in clinical texts.
  • Network Security: Investigating secure integration of AI systems in networks, focusing on ensuring data privacy and robustness against cyber threats.
  • Backend Development and Automation: Creating robust backend solutions, automating workflows to reduce manual effort, and integrating advanced AI capabilities into traditional software systems.
  • Web Scraping and Data Collection: Designing efficient web scraping tools to collect data for AI model training, while adhering to best practices in ethical data usage.
  • Integration of AI into Gaming Engines: Exploring the use of AI models in game development platforms like Unreal Engine and Unity to create smarter, adaptive gameplay and enhanced player experiences.

I am especially interested in bridging the gap between research and practical applications, whether it is through developing real-time AI systems for gaming, creating automation solutions that enhance productivity, or employing Generative AI to solve complex real-world challenges.