I am a data analyst and aspiring AI engineer passionate about turning data into insights and building machine learning solutions that create real business value.

After several years working in telecommunications and customer-facing roles, I have focused on analytics, automation, and data-driven decision-making using Python, SQL, Power BI, and Machine Learning.

Tech: Python • Pandas • Parquet • Bronze/Silver/Gold Architecture

A complete data engineering project that simulates real-time customer event streams.

The project continuously generates customer events (logins, purchases, cancellations), consumes them in a streaming-style architecture, and builds a complete Lakehouse pipeline consisting of Bronze, Silver, and Gold layers.

• Streaming simulation using JSON microbatches
• Automated ingestion into Bronze (Parquet)
• Silver: cleaned and standardized event table
• Gold: analytics-ready tables prepared for Power BI
• Modern data engineering architecture explained and implemented

Tech: FastAPI • Embeddings • Mock Mode • Retrieval Logic

An AI-powered support API combining retrieval, embeddings, and RAG-inspired reasoning.

The project includes a complete mock mode, allowing the entire system to be tested without any API costs, making it ideal for learning and portfolio development.

• FastAPI application with complete endpoint structure
• Embeddings (OpenAI or cost-free synthetic embeddings)
• RAG-style response generation based on relevant documents
• Structured response models with metadata and similarity scores
• Clean architecture and testability

Tech: Python • Streamlit • TF-IDF • Cosine Similarity

A FAQ chatbot built using NLP retrieval techniques that matches user questions against a knowledge base using TF-IDF vectorization and cosine similarity.

An interactive Streamlit interface makes it easy to test and adjust confidence thresholds and retrieval settings.

• TF-IDF + cosine similarity matching
• Adjustable confidence threshold and top-k retrieval
• Interactive Streamlit web application
• Easily extendable to embeddings and RAG architectures

Tech: Python • Scikit-Learn • Pandas • Seaborn

Customer segmentation using K-Means clustering to identify distinct customer groups based on demographics and purchasing behavior.

• K-Means clustering with StandardScaler
• Elbow Method for optimal cluster selection
• Cluster visualization and interpretation
• Insight-driven customer analysis

Tech: Scikit-Learn • XGBoost • Pandas

A complete churn prediction project comparing multiple machine learning models and evaluation techniques to identify customers at high risk of leaving.

• Logistic Regression, Random Forest, and XGBoost
• ROC-AUC, Precision, Recall, and F1 evaluation
• End-to-end preprocessing and modeling pipeline
• Analysis of churn drivers and customer behavior

Tech: Python • Scikit-Learn • XGBoost

A regression project focused on predicting house prices using a rich set of numerical and categorical features.

• Linear Regression, Random Forest, and XGBoost
• RMSE, MAE, and R² evaluation metrics
• Feature engineering and preprocessing
• Model explainability and interpretation

Tech: Power BI • DAX • Data Cleaning

An interactive Power BI dashboard analyzing Norway's aquaculture industry with a focus on regions, production volume, and long-term trends.

• Data cleaning and modeling
• DAX measures and KPI development
• Geographic visualizations
• Strong focus on storytelling and business insights

Python • SQL • Power BI • Scikit-Learn • Pandas • NumPy • Matplotlib • Seaborn • XGBoost

Git • GitHub • VS Code • Jupyter Notebook

Exploratory Data Analysis (EDA) • Feature Engineering • Classification • Regression • Clustering • Model Evaluation • Data Visualization

FastAPI • Streamlit • NLP • TF-IDF • Embeddings • Retrieval Systems • RAG-Inspired Architectures • API Development

• Learn Python 3
• Analyze Data with SQL
• Analyze Data with Microsoft Excel
• BI Dashboards with Power BI
• Data and Programming Foundations for AI
• Data Scientist: Analytics (Codecademy)

• Data Scientist: Machine Learning Specialist
• Build Chatbots with Python
• Creating AI Applications Using RAG
• Learn How to Build AI Agents

• Time Series Analysis and Forecasting
• AI Agents and LLM-Powered Applications
• Retrieval-Augmented Generation (RAG)
• MLOps and Production Machine Learning
• Data Engineering and Lakehouse Architectures
• Advanced Power BI and PL-300 Certification Preparation

I am building a portfolio that combines business understanding with technical expertise, focusing on demonstrating how data, machine learning, and AI can be transformed into practical insights and real-world business value.

I enjoy working at the intersection of analytics, software development, and product thinking, with a strong interest in AI-powered applications and data-driven decision-making.

💬 Open to collaboration, feedback, and professional discussions within Data, AI, Analytics, and Software Engineering.

GitHub: https://github.com/Runar-Olsen