Kimmo S.
Data Science And Machine Learning Expert With Outstanding Communication Skills And A Make-IT-Happen Attitude.
Experience
Jan 2025 - Dec 2025
1 yearPredictive maintenance in manufacturing industry
Upcognify Ltd
- Designed and trained deep neural networks to solve a predictive maintenance problem in the manufacturing industry
- Defined suitable performance metrics for the models and tracked them on an MLflow server set up for this purpose
- Coached the client’s managers and developers in machine learning, neural networks, and agile software development
- Technologies and methods: Python, NumPy, SciPy, Pandas, Scikit-learn, NLTK, Keras, PyTorch, TensorFlow, OpenCV, Detectron2, Meshroom, MLflow, Jupyter, Databricks, PySpark, AWS, Azure, GCP, SQL, Git, JavaScript, HTML, CSS, Java, Flask, Celery, MATLAB, R, C, C++
Jan 2024 - Dec 2025
2 yearsComputer vision for object detection
Upcognify Ltd
- Applied transfer learning to deep neural networks to detect new kinds of objects in industrial videos
- Created a photogrammetric 3D model of the scene to identify detected objects and preserve identities in new videos
- Technologies and methods: Python, NumPy, OpenCV, Detectron2, MLflow, Meshroom, Databricks, MS Azure, Jupyter Notebooks, Git, GitHub
Jan 2023 - Dec 2024
2 yearsAnomaly detection in aerial images and videos
Upcognify Ltd
- Wrote functions for anomaly detection in aerial images and videos where possible anomalies were highly ambiguous and training examples were scarce
- Solved the problem with unsupervised anomaly detection by comparing new images of the same scenery against existing ones in a statistical sense
- Technologies and methods: Python, NumPy, OpenCV, neural networks, Databricks, MS Azure, Jupyter Notebooks, Git, GitHub
Jan 2023 - Dec 2023
1 yearObject detection in industrial images
Upcognify Ltd
- Developed algorithms for object detection and segmentation in outdoor industrial images taken under varying conditions
- Created a custom web application to annotate a few hundred objects manually due to lack of annotated images
- Used transfer learning to train a neural network to solve the main detection problem
- Technologies and methods: Python, NumPy, OpenCV, neural networks, Databricks, MS Azure, Jupyter Notebooks, Flask, JavaScript, HTML, CSS, Git, GitHub
Jan 2022 - Dec 2023
2 yearsDigital image processing and analysis
Upcognify Ltd
- Designed and implemented a system that processes and analyses digital images from thermal sensors and photographs collected via IoT
- Implemented image compression, interpolation, registration, edge detection, distortion correction, artifact removal, and hot spot detection
- Built a web application with features such as image search, zoom, pan, selection, animation, and pixel highlighting
- Technologies and methods: Python, NumPy, OpenCV, Pillow, Databricks, MS Azure, Jupyter Notebooks, Flask, JavaScript, Plotly, Bootstrap, Tabulator, HTML, CSS, Git, GitHub
Jan 2021 - Dec 2022
2 yearsDeep learning for predictive maintenance
Upcognify Ltd
- Developed a deep learning anomaly detection system for predictive maintenance of industrial devices
- Used a network of IoT devices to record multiple signals and send them to a cloud platform for analysis
- Ran anomaly detection algorithms on acoustic emission data to spot abnormal activities in small unlabeled datasets
- Technologies and methods: Python, NumPy, SciPy, Pandas, Scikit-learn, TensorFlow, Keras, Plotly, Jupyter Notebooks, Databricks, PySpark, MS Azure, PowerBI, Git, GitHub
Jan 2021 - Dec 2022
2 yearsDeep learning for signal classification
Upcognify Ltd
- Developed software for deep learning–based signal classification including data exploration and visualization
- Performed digital signal processing to suppress noise and artifacts
- Trained and benchmarked multiple deep neural networks
- Technologies and methods: Python, NumPy, SciPy, Pandas, TensorFlow, Keras, MLflow, Plotly, Jupyter Notebooks, Git, GitHub, AWS SageMaker, AWS S3
Jan 2020 - Dec 2021
2 yearsNeural networks for time series analysis
Upcognify Ltd
- Conducted exploratory analysis of customer data stored in the client’s data lake
- Designed and implemented neural networks for time series analysis and customer support based on collected business cases and feedback
- Designed and implemented data pipelines and participated in founding a new data warehouse
- Technologies and methods: Python, NumPy, SciPy, Pandas, Scikit-learn, TensorFlow, Keras, Plotly, Jupyter Notebooks, Git, GitHub, AWS SageMaker, AWS S3, AWS Athena, AWS Glue, Jira, Asana, Confluence
Jan 2017 - Dec 2021
5 yearsMachine learning web app
Upcognify Ltd
- Designed and built a medical-grade web application that uses machine learning to analyze physiological signals
- Responsible for requirement analysis, architectural design, technology selections, hands-on implementation, testing, documentation, and maintenance
- Built data pipelines, implemented signal processing algorithms, trained and selected machine learning models, and implemented front-end and back-end modules
- Selected and managed cloud resources and advised technical and business personnel
- Technologies and methods: Python, NumPy, SciPy, Pandas, Scikit-learn, Keras, TensorFlow, Plotly, Jupyter Notebooks, Git, GitLab, AWS EC2, AWS S3, GCP, C, C++, DSP, MySQL, Eclipse, Flask, Celery, HTML, CSS, JavaScript, Scrum, Kanban, XP
Summary
Data science and machine learning expert with outstanding communication skills and a make-it-happen attitude. Three graduate degrees with excellent grades. Doctoral degree on the way. Years of experience in both the industry and the academia.
I provide full-stack data science services as a freelancer through my company Upcognify Ltd. My special areas of expertise are machine learning, computer vision, and time series analysis. I also have experience in natural language processing, cloud computing, optimization, full-stack web development, medical software, IoT, DevOps, and agile methods. I take great pride in delivering high-quality software that is easy to maintain and I can implement a solution in a proactive way from idea to production.
Languages
Finnish
NativeEnglish
AdvancedSwedish
IntermediateGerman
ElementaryEducation
Aalto University
Doctor of Science (Technology) · Computer Science · Finland
Aalto University
Master of Science (Technology) · Machine Learning and Data Mining · Finland
National Defence College
Military Officer (graduate degree) · Finland
...and 3 more
Certifications & licenses
Machine Learning With Tree-Based Models In Python
Object Detection In Industrial Images
Intermediate Python For Data Digital Image Processing And Analysis
Data Scientist With Python
Deep Learning With Python
Supervised Learning With Detectron2, MLflow, Scikit-Learn Notebooks, Git, GitHub
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.