Sigurd Roll Solberg

Sigurd Roll Solberg

Data Scientist | Founder | Nerd

About Me

I am a mathematician, AI researcher, and entrepreneur focused on bridging fundamental research with real-world application.

As the founder and CEO of Merlin, the team and I work to empower businesses to test, refine, and deploy their systems with greater accuracy and security with our AI solutions.

Machine Learning
95%
Data Analyst
90%
Data Visualization
85%
Big Data Technologies
80%

Technical Skills

My toolkit for solving problems and delivering impactful solutions.

Programming Languages

Python
Java
SQL
Bash

Machine Learning

PyTorch
Scikit-learn
Forecasting
Self-Supervised Learning
GNN
GenAI
Anomaly Detection
TDA

Data Visualization

Matplotlib
Seaborn
Plotly
Tableau
KeplerMapper

Big Data

Oracle
Clickhouse
Polars

Cloud and Platform

AWS
Google Cloud
Openshift
Kubernetes

Tools & Methodologies

Git
Docker
Jupyter
Agile
CI/CD
Pydantic
FastAPI
WandB

Experience

My professional journey and academic background.

Work Experience

Education

Atlas

A Large Financial Model

For the past two years, I have led the development of Atlas, a generational financial foundation model designed to transform how financial institutions detect fraud, assess risk, and safeguard financial systems. Built on a generalizable AI framework, Atlas unlocks broader applications in forecasting, anomaly detection, and systemic risk analysis.

Advanced Predictive Analytics

Uses temporal graph dynamics to contextualize vast amounts of transactional data.

Real-time Risk Assessment

Continuous monitoring and evaluation of financial risks across diverse portfolios

Transactions, The Language of Finance

Treats transactions as a semantic structure, encoding the flow of value much like language encodes meaning.

Python
PyTorch
Financial Modeling
Time Series Analysis
Temporal Graphs
Strategic Leadership

Merlin

Innovate with Synthetic Data

Merlin redefines how businesses harness data by generating high-fidelity synthetic datasets that retain the statistical integrity of real-world information while ensuring privacy, security, and fairness. Our approach enables seamless dataflows across complex systems, unlocking new possibilities in AI development, software testing, and financial modeling.

Privacy-Preserving Data

Generate synthetic data that maintains utility without compromising sensitive information

Bias Mitigation

Advanced algorithms to detect and reduce bias in training data

Scalable Data Generation

Cloud-based infrastructure capable of producing millions of synthetic records on demand

GANs
Differential Privacy
PyTorch
Cloud Computing
Data Ethics

Featured Projects

A selection of my most impactful data science projects.

Customer Segmentation Project

Customer Segmentation Model

Developed a clustering algorithm that identified 5 distinct customer segments, enabling targeted marketing campaigns that increased conversion rates by 18%.

Python
K-means
Tableau
View Project
Predictive Maintenance Project

Predictive Maintenance System

Built a time-series forecasting model that predicted equipment failures with 92% accuracy, reducing downtime by 35% and saving $1.2M annually.

TensorFlow
LSTM
IoT
View Project
NLP Sentiment Analysis Project

NLP Sentiment Analysis

Created a sentiment analysis tool that processed customer reviews in real-time, providing insights that improved product development and customer service.

BERT
NLP
AWS
View Project

Get In Touch

Interested in working together? Feel free to reach out through any of these channels.

Email

sigroll@gmail.com

LinkedIn

linkedin.com/in/sigurd-solberg/

GitHub

github.com/sigurdsolberg

Resume

Download PDF