Notebook
Dated notes from our research and development work.

- Why SI, not AI: the case for symbiotic intelligenceMost technology companies founded this decade put two letters in their pitch. We put two different ones in our name. What does Symbiotic Intelligence mean, and how does it shape what we do and why?
- What a brain-computer interface actually is: from EEG signal to decisionNo mind reading, no magic. A plain, accurate explanation of how microvolts on the scalp turn into a decision a machine can act on.
- The X principle: how we approach problems that seem impossibleThe X in our name stands for refusing the words “can’t be done”. Not by ignoring limits, but by turning them into questions. The four steps we follow when something does not work yet.
- Why I’m building a brain-computer interface in TurkeyA founder’s journey from Şanlıurfa to the frontlines of neural technology: why assistive communication must work in Turkish too, what building in an emerging market really takes, and why 2025 was the turning point.
- AI Regulation and EthicsFrom the EU AI Act to Türkiye’s data protection law, from bias testing to explainability: where AI regulation stands today, a framework for developing AI ethically, and a practical guide to how companies can adapt to the new rules.
- AI and Ethics: The Responsibilities of Data ScientistsData scientists play a key role in every stage of an AI system, from design to deployment. This article sets out their ethical responsibilities under six headings, from privacy and fairness to transparency and sustainability.
- Python for Data Analysis: Machine Learning with Scikit-LearnA hands-on introduction to machine learning with Scikit-Learn: supervised and unsupervised learning, hyperparameter tuning, cross-validation, feature selection, model interpretation, ensembles, metrics, pipelines and a case study on the Iris data set.
- Python for Data Analysis: Matplotlib and Data VisualizationMatplotlib is one of Python's most widely used plotting libraries. This guide covers line, bar, scatter, histogram and pie charts, then customization, subplots, logarithmic and 3D plots, with code for each.
- Data Analysis with Pandas: Manipulating and Visualizing DataPandas is one of the most useful Python libraries for anyone who works with data. This guide walks through DataFrames and Series, loading, cleaning, summarizing, grouping, plotting, time series and merging, then puts it all together on a synthetic data set.
- In the light of the Renaissance and the wisdom of Tibetan monksA personal note on taking part in NUWA NFTs’ Dejavus series: five words (initiation, time, spiritual, mind, flow) turned into an artwork by AI artist Çağla Öztürk and music by E. Buğra Balcı.
- What Is an Imbalanced Data Set, and How Do You Deal with It?Your model reports 99.9% accuracy and still misses the fraud. Using a credit card fraud data set, we show why imbalanced data misleads, which metrics to look at, and how to rebalance with oversampling, SMOTE and undersampling.

