Welcome to
Alfred
Vikström's
Portfolio
About Me
My name is Alfred Vikström and I love making an impact. I believe that choosing curiosity over fear is a better way to live, and this mindset has shaped my approach to problem-solving and innovation. One of my greatest strengths has always been my ability to ask the right questions. I thrive on understanding the complexities of systems, designing and prototyping ideas, identifying bottlenecks, and finding ways to optimize and improve them.
I like to work on problems on the edge of my knowledge and continuously learn more while making a difference. I believe AI is our time's most influential technology, and I want to be part of focusing its potential in a suitable direction.
Skills
OPAL (ORU Pedagogic Assistance for Lecturers)
AI-Powered Assistant for Higher Education
Initiator, Strategic Lead & AI Developer
2025–
OPAL began as a student project and evolved into a university-sponsored AI innovation venture. I initiated and led the development of an AI assistant aimed at improving pedagogical quality in higher education. By leveraging lean strategy, design thinking, and effectuation, I guided the project from vague opportunity to a concrete MVP adopted for institutional testing. Today, OPAL continues as a strategic partnership with Örebro University and AI Impact Lab.
- Initiation & Strategic Framing: Identified the opportunity to improve higher education using AI. Framed the problem space, defined a vision and pivoted the solution from student-focused to lecturer-focused to align with institutional capabilities and resources.
- Stakeholder Engagement: Led human-centered outreach across university stakeholders, including pedagogy departments, innovation hubs, and students, to validate needs and establish collaboration pathways.
- AI Prototyping & Development: Built the OPAL MVP using tools like Langflow, Python, and Supabase. Integrated LLMs into a web-based chatbot platform for assisting educators with course planning and pedagogical activity design.
- Team Leadership & Process Design: Established a transparent, trust-based team culture. Applied effectuation, agile, and lean startup principles to guide decision-making, delegate responsibilities, and foster open collaboration.
- Institutional Integration & Future Strategy: Negotiated university partnerships, established OPAL as a candidate for institutional adoption, and secured funding through the Entrepreneurs in Residence (EIR) program for summer 2025 business development.
Prince
Stock Analysis Algorithm
Technical Lead and Developer
2021-
Prince is a personal project focused on creating an advanced stock analysis tool utilizing custom algorithms and deep learning models for technical analysis of stock data. The project, started during my time at the AI Business Consulting program, quickly grew to include sophisticated data pipelines and signal analysis.
- Algorithm Development: Led the design and development of a custom algorithm for efficient stock comparison using historical price data.
- Deep Learning Integration: Implemented GRUs, LSTMs, CNNs, and Transformers to enhance stock prediction accuracy and trend analysis.
- Data Pipeline Optimization: Optimized data pipelines and calculations by using techniques like vectorization in NumPy, multi-processing and hyper parameter optimization.
- Collaboration: Worked alongside a financial expert and a PhD data scientist to integrate insights from both fields into the technical strategy.
Smidigt Handel
Business Process Optimization
AI Consultant
2024-
Smidigt Handel is a Swedish company focused on buying, refurbishing, and selling used electronics. As a consultant, I help identify and eliminate bottlenecks in their workflow and data pipeline. By automating manual processes and integrating AI-driven solutions, I optimize their operations for better efficiency and scalability. The project is still in early stages, but there are already noticeable improvements in time, resource, and cost management.
- Bottleneck Identification & Prototyping: Identified workflow bottlenecks and built Python-based prototypes to automate and streamline processes.
- Business Process Automation: Developed RPA solutions and custom Python scripts to automate manual tasks, improving efficiency.
Neurodiverse Potential
Online Platform Development
Product Owner
2024
Neurodiverse Potential provides assessments for children and young people with neurodiverse challenges (such as ADHD and dyslexia) to help them receive support in schools. I served as the product owner for the development of an online platform that facilitates remote assessments, featuring multiple tests, a result visualization dashboard, and GDPR compliance. The platform is fully functional in its beta phase, with core features in place, and is being used by the company with plans for further test integrations.
- Product Ownership: Managed the scope and prioritized features, ensuring alignment between client requirements and the development team.
- Test Standardization & UX Challenges: Ensured that online tests met strict standards set by psychology organizations, enabling their use in educational settings.
- GDPR Compliance: Oversaw the implementation of GDPR compliance to ensure the secure handling of user data in accordance with legal regulations.
- Dashboard & Results Page Development: Contributed to the design and functionality of the platform's result visualization and dashboard, improving user and assessor experience.
- Platform Deployment: Led the team through the platform's beta launch, with core features in place, and supported plans for future test integrations.
CastWise
AI-Driven Podcast Post-Production Workflow
AI Workflow Developer / Producer
2023-
CastWise is an AI-driven tool developed to streamline the post-production workflow for a podcast I produce at Savantic, a Stockholm based AI Consulting firm. The tool uses generative AI and large language models (LLMs) to automate transcription, correct errors, and assist with content creation, such as generating marketing reels and summaries. By integrating these features into video editing software, CastWise enhances productivity by marking key moments in the podcast for easier post-production.
- LLM Pipeline Development: Developed an AI-driven post-production tool using LLMs to automate transcription, error correction, and content creation tasks.
- Prompt Engineering: Designed and optimized prompts for specific tasks, improving transcription accuracy and generating summaries for marketing.
- Learning & Exploration: Gained hands-on experience working with LLM API's, how to communicate with them and exploring the possibilities and potential of this powerful technology.
IKEA Sentiment Analysis Demo
AI-Driven Sentiment Analysis for Customer Feedback
Technical Lead
2021 (Part of Education)
As part of the AI Business Consulting program at Hyper Island, I worked on a project with IKEA to develop a solution for gathering and analyzing customer feedback. I led the development of a sentiment analysis demo using a BERT model to classify customer comments as positive, negative, or questions. Data was scraped from YouTube videos featuring IKEA experiences, and the model was trained to provide insights on customer sentiment. The solution was presented to IKEA, who gave positive feedback and expressed interest in integrating the technology.
- Technical Lead: Led the development of a BERT-based sentiment analysis model for classifying customer feedback.
- Data Scraping & Preparation: Built a data scraper to collect YouTube comments about IKEA and prepared the data for model training.
- Cross-Functional Teamwork: Collaborated with a cross-functional team of business and design students to align the technical solution with business goals.
Action for Society – Safety Monitoring
AI-Driven Safety Monitoring
Technical Lead / Data Strategist
2022 (Part of Education)
The Action for Society project aimed to monitor the feeling of safety in urban areas, using localized crime data to overcome the limitations of low-resolution survey data. I led the development of a machine learning model that predicted safety scores based on 10 years of crime data across Sweden, providing actionable insights for smaller neighborhoods. This strategy helped the team enhance safety monitoring in urban areas, with the model successfully predicting safety scores for areas without direct survey data.
- Strategic Data Solution: Devised a strategy to use localized crime data to enhance the resolution of safety monitoring in urban areas.
- Machine Learning Model Development: Led the development of a machine learning model to predict safety scores based on crime data and survey data.
- Data Acquisition & Preparation: Acquired and processed 10 years of crime data from various sources in Sweden to provide highly localized data for the model.
- Cross-Functional Teamwork: Collaborated with a cross-functional team to align technical solutions with societal safety monitoring goals.
Personal Portfolio Website
Dynamic and Responsive Portfolio Website
Web Developer / Designer
Ongoing
I developed my personal portfolio website from scratch to improve my web development skills and showcase my projects. Using generative AI-driven coding tools for learning, I built a fully responsive website with a custom JavaScript animation mimicking a lava lamp as the dynamic background. The site was developed using modern frameworks like React, Shadcn, and Tailwind CSS, and focuses on providing a visually appealing and smooth user experience. This project helped me deepen my understanding of web development and UX design.
- Custom Lava Lamp Animation: Developed a custom JavaScript animation mimicking a lava lamp, used as a dynamic background for the website.
- Responsive Web Design: Built a fully responsive website using React and Tailwind CSS, ensuring seamless performance across devices.
- Learning Through AI Tools: Utilized generative AI coding assistants to aid in development and learn new frameworks and coding techniques.
- UX Design: Focused on designing a smooth and visually appealing user experience that effectively showcases my portfolio.