Moritz
Exploring Innovation in Data Analytics & AI Systems
21 | Artificial Intelligence | Data Science | Software Engineering
About
My Introduction
Hooked on AI
I'm deep in the world of AI, exploring its transformative potential across cutting-edge technologies. Beyond AI, I dive into massive software engineering projects that scale complex systems and push architectural boundaries.
Passion for Innovation
I'm passionate about Machine Learning-transforming abstract code into models that recognize images or understand context feels like magic.
Breaking Barriers
I thrive on solving complex problems and turning ideas into tech that makes a difference—those breakthrough moments are what drive me.
Skills
My technical & miscellaneous skillsAI & Data Science
3+ Years XPComputer Vision
92%Natural Language Processing
85%Generative Ai
90%Probability & Statistics
85%Data Analytics & Visualization
80%Mlops
90%Distributed DNN Training
85%Multi Agent Models
85%Programming
3+ Years XPPython
95%C & C++
90%C#
90%SQL
95%Computing
2+ Years XPDocker
90%Microsoft Azure
80%Amazon Web Services
75%Virtualization
75%HTML
90%CSS
85%JavaScript
75%Angular
85%React Native
85%BackEnd
2+ Years XPPython - Flask, Fast API
95%Node JS
80%Misc
2+ Years XPGit
90%Linux
90%DevOps
75%Software Development
85%Leadership
95%Communication
90%Time Management
98%Experience
My journey in the academic & professional front
Data Science Engineer
Eviden AustriaData Science Engineer
Smart DigitalFreelancer
Paramedic
Portfolio
My works, projects & contributions
Smart Digital - Geospatial Object Detection For
Railways
Utilizes computer vision and instance segmentation through a dual monitoring system of mounted track cameras and autonomous drones. The system continuously scans tracks for damage, misalignment, or wear while also detecting and mapping railway assets like signals, signs, and switches. All track conditions and infrastructure objects are accurately mapped to their precise geospatial positions on a secure monitoring interface.
View
MarketVision AI - Reinforcement Learning for Financial Trading
Utilizes Deep Q-Networks (DQN) to analyze candlestick chart images, enabling the system to predict market trends and make trading decisions. After rigorous testing, the model demonstrates the ability to slightly outperform market benchmarks, showcasing its potential for enhanced financial forecasting.
View
UMLify - Automated UML Diagram Generation from
Code
Leverages static code analysis to automatically generate UML diagrams from uploaded source code. This tool streamlines software documentation, aiding developers in visualizing system architecture and improving team collaboration.
View
AgeLens AI - Image-Based Age Prediction
Employs advanced deep learning models to analyze facial images and accurately predict the age of individuals. The system leverages computer vision techniques to deliver reliable age estimations, with potential applications in identity verification, personalized marketing, and social analytics.
View
TourPlanner - Interactive Map-Based Itinerary Builder
Enables users to design personalized tours by selecting destinations, activities, and routes displayed on an interactive map. The system provides real-time suggestions and visualizes the itinerary, making trip planning intuitive and engaging.
View
Eviden - Video-Based Pipe Damage Detection
Integrates template matching to analyze video frames and identify regions of interest, which are then processed through a Convolutional Neural Network (CNN) to detect pipe damage. This system ensures efficient and accurate pipeline monitoring, enabling timely maintenance and reducing operational risks.
Closed Source
PaperLess - Distributed Searchable Document
System
A scalable and distributed document management system built with independent Docker containers. It enables seamless indexing and searching of documents, ensuring high availability and flexibility for deployment across multiple environments. Perfect for organizations requiring efficient and distributed document retrieval.
View
Multi-Agent Camel Detection and Ranking System
A specialized program that detects camels based on the training set. Another model then checks if the camel is black. If all models pass, another model ranks the camel based on a dataset which has the price of camels at auction. Some features are e.g size, strength.. The system automates the process of camel auctions, improving efficiency and fairness.
View
Eviden - Cable Length detection
A program analyzes an image of a cable car cable. It calculates the length of the cable shown in the image. The program uses image processing and math for the calculation. It then displays the cable length as the result.
Closed Source
Style Transfer
A style-transfer application that can apply art styles or textures over new images.
View
GeoAI - Image-Based Country Detection
Employs advanced computer vision and machine learning models to analyze images and accurately identify their country of origin. The system leverages distinctive geographical, cultural, and architectural features, offering applications in geolocation, digital forensics, and tourism analytics.
View
RAG System - Retrieval-Augmented Generation
Uses word embeddings to identify context and then leverages large language models (LLMs) for accurate and minimal-hallucination responses. Ideal for precise, context-aware content generation.
Closed Source
VoiceGuard - Intelligent Trigger Word Detection
A Neural Network that detects the trigger word "activate". The system uses a combination of background audio, random word snippets, and "activate" snippets for training. When it hears "activate", it produces a chiming sound. The training data includes audio from various environments for robust detection.
View
JazzGenius - LSTM-Powered Music Generation
An LSTM network trained on a corpus of Jazz music to generate original compositions. The system predicts the next sound in a sequence, creating a chain of musical notes. Post-processing ensures the output adheres to musical conventions, producing coherent and pleasing Jazz melodies.
View