Maximilian Unfried

Rising Codementor

US$0.00

For every 15 mins

ABOUT ME

Curious data scientist, AI practioner and entrepreneur

I am a business minded data scientist with an entrepreneurial spirit. My experience is in hands-on product development and implementation of data driven products in the Natural Language Processing and Computer Vision domain using state of the art machine learning algorithms and technologies. I have experience working in an international environment and leading cross-cultural teams, and the ability to communicate fluently in German and English with team members, customers and stakeholders.

German, English

Berlin (+01:00)

Joined October 2018

EXPERTISE

Deep Learning

4 years experience

My Deep learning experience is of theoretical and practical nature. My mathematical foundations were build in university and strengthend ...

My Deep learning experience is of theoretical and practical nature. My mathematical foundations were build in university and strengthend by implementing algorithms from scratch, before Deep Learning was a hot topic. Through hands-on experiencen in industrial settings I acquired the skills needed to build robust models, that perform on messy real-world data. My skillset ranges from data acquisition, data labeling, data cleaning, developing network architectures over to training, testing, finetuning and continously improving Deep Learning models.

Artificial Neural Networks Deep Learning Python Keras Computer Vision Transfer learning Machine learning TensorFlow

Computer Vision

3 years experience

Throughout my career I've been involved in Computer Vision research and building Computer Vision applications and products. I am profici...

Throughout my career I've been involved in Computer Vision research and building Computer Vision applications and products. I am proficient in state of the art Computer Vision techniques using Deep Learning for Image Classification, Image Clustering and Object Detection. In addition I have significant knowledge of Data Acquisition strategies and best practices for Data Labeling, which is key as your model only performs as good as your data is. As the saying goes: Garbage in, Garbage out.

Deep Learning Python Transfer learning

NLP

2 years experience

My experience involves building NLP systems for social media analytics. I've been working on projects involving Text Classification, Text...

My experience involves building NLP systems for social media analytics. I've been working on projects involving Text Classification, Text Clustering, Language Detection and Sentiment Analysis, using traditional feature engieering and preprocessing as well as state of the art methodologies like word embeddings.

Python PandasWordembeddings

Machine learning

5 years experience

Machine Learning has been my companion for the last couple of years. I've been applying the most common traditional algorithms like Naive...

Machine Learning has been my companion for the last couple of years. I've been applying the most common traditional algorithms like Naive Bayes, Nearest Neighbours, SVMs, Perceptrons, k-means etc. to various types of problems and on data including language, images, genetics, ecommerce etc. I am well versed in developing models and measure their performance, making sure that they are not biased.

Machine learning Computer Vision Deep Learning

Machine learning Computer Vision Deep Learning NLP Keras TensorFlow Pandas

Data Science

3 years experience

I have thorough experience in analysing lage amounts of data and extracting significant insights and putting together reports. In additi...

I have thorough experience in analysing lage amounts of data and extracting significant insights and putting together reports. In addition to my analytical skillsset I am good at reporting, presenting and able to tell a narrative about the data, so that relevant stakeholders can understand insights and relevance to the business.

Python Pandas NumPy

Python Pandas NumPy Matplotlib Data analysis Data analyticsData science in python

Data analytics

7 years experience

Having a STEM education in the disciplines of Biophysics and Artificial Intelligence and practicing AI in the industry afterwards my prof...

Having a STEM education in the disciplines of Biophysics and Artificial Intelligence and practicing AI in the industry afterwards my professional career revolves around data analytics. Over the years I have been analysing, visualizing and extracting insights from biological, financial, economical, text and image data sources.

Data analytics Data analysis Data Science

Data analytics Data analysis Data Science Pandas PythonData science in python

REVIEWS FROM CLIENTS

Maximilian's profile has been carefully vetted and approved as a Codementor. Connect with Maximilian now, and leave a review for them once you're done!

EMPLOYMENTS

Chief AI Officer

deepgreen

2017-07-01-Present

As Chief Artificial Intelligence Officer I am in charge of the internal and external AI strategy including data acquisition and data l...

As Chief Artificial Intelligence Officer I am in charge of the internal and external AI strategy including data acquisition and data labeling operations, as well as deep learning model development. Furthermore, I was involved in the hiring process, leading the engineering team, fundraising and business development. My main work has been on pest and disease detection on RGB images of plants. TensorFlow and Keras have been used to build Convolutional Neural Networks based feature extractors for object detection using Faster RCNN and RetinaNet inspired architectures, which resulted in the first commercial usage of the disease detection service. For the purpose of improving the data I also developed tools for EDA, visualization and data cleaning in Python.

Python

Machine learning

Computer Vision

Python

Machine learning

Computer Vision

Data analysis

Data Science

Google Cloud Platform

Deep Learning

Data analytics

Keras

Data Scientist

Sentifi

2016-07-01-2017-08-01

At the award winning FinTech startup Sentifi I used Natural Language Processing and Machine Learning algorithms such as Naïve Bayes, ...

At the award winning FinTech startup Sentifi I used Natural Language Processing and Machine Learning algorithms such as Naïve Bayes, k-Means, TFIDF and Word Embeddings to build models for classification and clustering of financial events from tweets and newspapers. These models are used by the world’s leading financial services organizations to get unique insights in over 50.000 traded companies, currencies and commodities and the events that impact them and help them to make better investment decisions.

Python

SQL

MongoDB

Python

SQL

MongoDB

Pandas

Machine learning

Data analysis

Data Science

Data analytics

Data science in python

Guest Researcher

NCTU Machine Learning Lab

2015-08-01-2015-12-01

As a guest researcher at the NCTU Machine Learning Lab, I conducted research on machine learning algorithms for sparse classification wi...

As a guest researcher at the NCTU Machine Learning Lab, I conducted research on machine learning algorithms for sparse classification with application to “Automated Facial Expression Recognition”.

Machine learning

Computer Vision

MATLAB

Machine learning

Computer Vision

MATLAB

PROJECTS

Object Detection for Plant Diseases

2018

This project involved manual data acquisition, data labeling and model development. Data has been collected on site on farms, using high...

This project involved manual data acquisition, data labeling and model development. Data has been collected on site on farms, using high resolution imagery and on the spot labeling to guarantee highest labeling quality. A main component was to develope a non-biased dataset that accomodated for all the variables in crops like lighting, grow cycle stage, disease distribution etc. so that it resembled the real world environment. It also involved an advanced labeling strategy that was specific for plant diseases, to guarantee the highest quality of labels. After enough clean data was acquired a RetinaNet inspired object detector was trained with a ResNet-50 backbone yielding a production ready model.

Machine learning

Computer Vision

Data analysis

Machine learning

Computer Vision

Data analysis

Big Data

Data Science

Deep Learning

Data analytics

Data science in python

Weed detection in Alfalfa fields

2018

Using a Faster RCNN inspired 2 stage detector architecture with a ResNet-100 model as a backbone and transfer learning to detect differen...

Using a Faster RCNN inspired 2 stage detector architecture with a ResNet-100 model as a backbone and transfer learning to detect different kind of weeds in outdoor alfalfa fields. Production model reached an mAP of 0.78.

Computer Vision

Data analysis

Deep Learning

Computer Vision

Data analysis

Deep Learning

Data analytics

Data science in python