Neven Pičuljan, Deep Learning Developer in Zagreb, Croatia
Neven Pičuljan

Deep Learning Developer in Zagreb, Croatia

Member since September 19, 2016
Neven is a passionate deep learning/machine learning research and development engineer with six years of experience. He has extensive experience working with cutting-edge technologies and a strong ability to understand and solve problems efficiently. He communicates extremely well and is looking for additional freelance projects to challenge himself with.
Neven is now available for hire

Portfolio

Experience

Location

Zagreb, Croatia

Availability

Part-time

Preferred Environment

Git, CLion, PyCharm, Linux

The most amazing...

...thing I've built is a face recognition system; I scraped the data set online, trained the model in Torch, and wrote the neural network inference engine in C.

Employment

  • CEO

    2018 - PRESENT
    Pičuljan Technologies, Artificial Intelligence Research and Development and Consulting
    • Created models for time series analysis, computer vision, and NLP.
    • Built an AI library and associated products.
    Technologies: AWS, Flask, SQLAlchemy, PostgreSQL, Git, Apache Kafka, Docker, C, C++, OpenCV, PyTorch, Python
  • Artificial Intelligence Specialist

    2017 - PRESENT
    Toptal Clients
    • Worked on various AI projects (computer vision, time series analysis, NLP, etc.).
    • Implemented computer vision algorithms.
    • Worked with time series data.
    • Implemented a server for AI models.
    • Implemented a data visualization web application.
    Technologies: OpenCV, TensorFlow, Caffe, PyTorch, Python
  • Machine Learning Engineer

    2020 - 2020
    NDA (via Toptal)
    • Worked on a text clustering algorithm for an eCommerce project.
    • Contributed to the generation of synthetic text data for training text embedding extractors.
    • Worked on training and evaluating a text embedding extractor.
    • Helped reduce the dimensionality of text embeddings and visualization of text embedding clusters.
    Technologies: Python, Scikit-learn, PyTorch, Plotly, Matplotlib, SpaCy
  • AI Consultant

    2020 - 2020
    NDA (via Toptal)
    • Consulted for the client on how to create, improve, and deploy an image similarity model.
    • Created a baseline system to perform image similarity estimation.
    Technologies: Scikit-learn, Pandas, SciPy, NumPy, PyTorch, Python
  • AI Developer

    2020 - 2020
    NDA (Fintech Client; via Toptal)
    • Trained multiple time series analysis models for predicting price behavior in the future.
    • Deployed multiple time series analysis models.
    • Integrated several different finance APIs.
    Technologies: Quandl API, Google Cloud Platform (GCP), AWS, Scikit-learn, Pandas, SciPy, NumPy, Theano, TensorFlow, Keras, Python
  • Computer Vision Developer

    2018 - 2019
    NDA (Healthtech Client; via Toptal)
    • Trained multiple computer vision models for classification, segmentation, 3D reconstruction, and more.
    • Deployed multiple computer vision models.
    • Organized the protocol for data collection and annotation.
    Technologies: Google Cloud Platform (GCP), AWS, Scikit-learn, Pandas, SciPy, NumPy, Open Neural Network Exchange (ONNX), Core ML, OpenCV, Scala, PyTorch, Python
  • ML/AI Consultant

    2017 - 2019
    Precious
    • Trained different computer vision models for detection, recognition and clustering.
    • Deployed different computer vision models for iOS using CoreML and ONNX.
    • Worked on the protocol for data collection and annotation.
    Technologies: AWS, Core ML, Open Neural Network Exchange (ONNX), Scikit-learn, Pandas, SciPy, NumPy, OpenCV, TensorFlow, PyTorch, Python
  • Co-founder/AI Engineer

    2017 - 2019
    Poze
    • Created a neural network inference engine for Android.
    • Trained a pose estimation model.
    • Created a testing framework for the pose estimation model.
    • Created a pose estimation library in C/C++.
    Technologies: C, C++, OpenCV, TensorFlow, Python
  • Developer

    2018 - 2018
    Fitz-Gerald Research Publications
    • Worked on a web-based application for screening time series data using proprietary algorithms.
    Technologies: AWS, SQLAlchemy, Dash, Flask, Scikit-learn, Pandas, SciPy, NumPy, Python
  • ML Engineer

    2018 - 2018
    NDA (via Toptal)
    • Created an image/text classifier using PyTorch and a large database.
    • Deployed an image/text classifier on AWS.
    • Created a user interface using Dash by Plotly.
    Technologies: Dash, AWS, Scikit-learn, Pandas, SciPy, NumPy, Flask, PyTorch, Python
  • ML Engineer

    2017 - 2017
    NDA (via Toptal)
    • Trained neural networks for image similarity.
    • Deployed neural networks for image similarity as a web service.
    • Created a protocol for data collection and annotation.
    Technologies: Dash, AWS, Scikit-learn, Pandas, SciPy, NumPy, Flask, PyTorch, Python
  • Python Django Developer

    2017 - 2017
    NDA (via Toptal)
    • Worked on a web-shop-like web application.
    Technologies: REST API, Heroku, PostgreSQL, Django, Python
  • Research Engineer

    2016 - 2017
    Visage Technologies
    • Collected the data set for building a face recognition system.
    • Built a training tool and trained a face recognition neural network model using Torch and TensorFlow.
    • Created a testing framework.
    • Coded the neural network inference engine in C/C++.
    • Cross-compiled the neural network inference engine.
    Technologies: Android Studio, Robot Operating System, OpenCV, CLion, PyCharm, Microsoft Visual Studio, BLAS, C++, C, TensorFlow, Torch, PyTorch, Linux
  • Django Developer

    2015 - 2015
    Mobilne Aplikacije d.o.o.
    • Developed Django applications and REST web services.
    • Created database models.
    • Scraped data from the internet.
    Technologies: MySQL, PyCharm, Django REST Framework, Linux, Python, Django
  • Machine Learning/Data Mining Intern

    2015 - 2015
    Bisnode
    • Collected data to create a named entity recognizer for the Croatian language.
    • Trained a named entity recognizer for the Croatian language.
    • Created a testing framework.
    • Made a web service to expose the named entity recognizer.
    • Crawled various types of data from the internet.
    Technologies: C++, C, Python, Linux
  • Software Engineering Intern

    2014 - 2014
    Visage Technologies
    • Developed a video face annotator.
    • Created tests for the face annotator.
    • Created a user’s manual for the face annotator.
    Technologies: Microsoft Visual Studio, OpenCV, C, C++, Windows, Linux
  • Teaching Assistant on Probability and Statistics

    2014 - 2014
    University of Zagreb, Faculty of Electrical Engineering and Computing
    • Prepared students for the exams.
    • Created assignments for the students.
    • Corrected students' exams.

Experience

  • Deep Visual Biometrics (Development)
    http://www.visualsweden.se/aktuella-projekt/forstudie-deep-visual-biometrics/

    I created a feasibility study and demo for a face recognition system that I developed at Visage Technologies. The demo was written in C/C++ and Python. I collaborated with the Swedish Police, the Swedish National Forensic Centre, and the Swedish Defence Research Agency on this project.

  • Neural Network for Function approximation Using Levenberg-Marquardt Algorithm in Torch Framework (Development)

    A neural network for function approximation using the Levenberg-Marquardt algorithm. I tested the code on various functions and used Torch framework and Python.

  • Credit Card Application Classifier (Development)

    A simple classifier in R for credit card applications. I was given a data set with users’ interactions and experimented with various machine learning algorithms: SVMs, decision trees, random forests, logistic regression, etc.

  • Clustering (Development)

    A project in data mining. I was given a data set with detailed information about interactions of visitors with different stations at Copernicus Science Centre in Warsaw, Poland. The goal was to characterize the flow of visitors through these stations and to segment visitors into separate categories/segments. I used R.

  • Contour Detection (Development)

    A system for detection and localization of a 2D contour (human head) in an image, where many such contours of different size could exist. For this purpose, I applied the generalized Hough Transform (GHT). The system was written in Python.

  • Operations on Graphs in LISP (Development)

    An implementation of various operations on graphs in LISP: finding cycles in graphs, finding paths from one node to another in graphs, checking if the binary tree is symmetric, depth-first order graph traversal, finding maximum depth of a binary tree, and finding a leaf with a maximum value in a binary tree.

  • Expert System in Prolog (Development)

    An expert system created in Prolog for animal identification.

  • Face Recognition (Development)

    A face recognition system. It was trained and tested in Torch framework. The data set was made of publicly available data sets.

  • Deep Regression for Face Alignment (Development)

    Research conducted on different algorithms for face alignment.

  • Answer Selection in Community Question Answering (Development)

    A system to automate the classification of Stack Overflow's posts in the answer thread into three categories: One for those that answer the question well. Another for those that can be potentially useful to the user (e.g., because they can help educate him/her on the subject). Lastly, group those that are just bad or useless.

    I experimented with various machine learning algorithms (scikit-learn): Gaussian naive Bayes, SVMs, and random forests.

  • Pedestrian Detection in Urban Environments Using Detectors Based on Contours (Development)

    A system to do pedestrian detection in urban environments using contour based detection. It was written in Python using Numpy, Scikit-learn, and OpenCV.

  • Performance-driven Animation as a Web Application (Development)

    Performance-driven animation as a web application. Face tracking was used to gain motion in the face of an animated virtual character. The graphics system used to build the application was Three.js based on WebGL. The face-tracking system used to build the application was Visage|SDK.

  • SkyRail Computer Game Controlled with Head Movements (Development)
    https://www.youtube.com/watch?v=QrsVpX5-LXo

    A computer game controlled with head movements. It was written in C# using Unity game engine.

  • Schooling Flappy Bird: A Reinforcement Learning Tutorial (Publication)
    Leveraging DeepMind's breakthrough AI approaches takes some work, but the results are astounding. In this article, Toptal Freelance Deep Learning Engineer Neven Pičuljan guides us through the building blocks of reinforcement learning, training a neural network to play Flappy Bird using the PyTorch framework.
  • Introduction to Deep Learning Trading in Hedge Funds (Publication)
    In this article, Toptal Freelance Software Engineer Neven Pičuljan introduces you to the intricacies of deep learning in hedge funds and finance in general.

Skills

  • Languages

    C++, C, Python, R, Lisp, Bash, Prolog, JavaScript, Perl, Java, Scala
  • Frameworks

    Core ML, Django REST Framework, Django, Caffe, Flask, Robot Operating System
  • Libraries/APIs

    LSTM, BLAS, TensorFlow, OpenCV, PyTorch, Stanford NLP, Quandl API, SQLAlchemy, NumPy, SciPy, Pandas, Scikit-learn, REST API, Keras, Theano
  • Tools

    NER, AWS CLI, Microsoft Visual Studio, PyCharm, Android Studio, CLion, Stanford NER, AWS ECS, Subversion (SVN), Git
  • Paradigms

    Data Science
  • Platforms

    AWS EC2, Linux, Android, Windows, Google Cloud Platform (GCP), Docker, Heroku, Apache Kafka
  • Storage

    AWS S3, PostgreSQL, MongoDB, MySQL
  • Other

    Sentiment Analysis, Probability Theory, LSTM Networks, Gated Recurrent Unit (GRU), SVMs, Support Vector Machines (SVM), Random Forests, Decision Trees, Decision Tree Classification, Decision Tree Regression, Logistic Regression, Linear Regression, Classification, Text Classification, Text Analytics, Computer Vision Algorithms, Statistics, Recurrent Neural Networks, Natural Language Processing (NLP), Deep Neural Networks, Deep Reinforcement Learning, Reinforcement Learning, Artificial Intelligence (AI), Computer Vision, Deep Learning, Machine Learning, Torch, AWS, Dash, Open Neural Network Exchange (ONNX)

Education

  • Master's degree in Computer Science
    2015 - 2016
    Warsaw University of Technology - Warsaw, Poland
  • Master's degree in Computer Science
    2014 - 2016
    University of Zagreb, Faculty of Electrical Engineering and Computing - Zagreb, Croatia
  • Bachelor's degree in Computer Science
    2011 - 2014
    University of Zagreb, Faculty of Electrical Engineering and Computing - Zagreb, Croatia

Certifications

  • Convolutional Neural Networks
    NOVEMBER 2018 - PRESENT
    Coursera
  • Deep Learning Specialization
    NOVEMBER 2018 - PRESENT
    Coursera
  • Sequence Models
    NOVEMBER 2018 - PRESENT
    Coursera
  • Structuring Machine Learning Projects
    NOVEMBER 2018 - PRESENT
    Coursera
  • Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
    OCTOBER 2018 - PRESENT
    Coursera
  • Neural Networks and Deep Learning
    OCTOBER 2018 - PRESENT
    Coursera
  • Artificial Intelligence
    APRIL 2018 - PRESENT
    Toptal, LLC
  • Data Science
    APRIL 2018 - PRESENT
    Toptal, LLC
  • Machine Learning
    SEPTEMBER 2014 - PRESENT
    Coursera

To view more profiles

Join Toptal
Share it with others