01 / Live product
Nordic Trails
A hiking-planning platform shaped around routes, gear and the practical details of a trip.
Product direction, functionality & design · Claude Code build
The product and my role →Data science, predictive modelling & applied AI · Copenhagen
Understand the data.
Build something useful.
Exploring patterns, evaluating predictions and building useful data tools. Independent work across quantitative analysis, forecasting and document AI, grounded in enterprise delivery experience.
Selective projects · Scope and timing by agreement
11+ years in enterprise deliverySiemens · ZEISS · TÜV SÜD Prior employmentIndependent practice · Copenhagen
From idea to working tool
Selected work.
Independent products and experiments, with a clear account of my role.
When the benchmark gets stricter
Historical research · 2020 start · methodology in the notes02 / Research tool
Dividend dashboard
An original Python dashboard expanded with Claude into a research system for portfolio backtesting, benchmark comparisons and paper trading.
Original programming · Claude-assisted benchmark engine
Read the research notes →03 / Local AI experiment
Archat
A small local AI assistant for Danish building specifications, exploring document search, chat and Word-template drafting.
Small MVP · Qwen3.5 4B · Jetson Orin Nano
Explore the MVP →04 / Internal tool
JA Tracker
A personal tool for planning work, study, projects and available time.
Project manager, customer & product owner · Claude-built
See the workflow →How we start
One question.
A clear next step.
Start with a question about your data, a prediction to evaluate or a document-heavy task. A small assessment can establish what is feasible and what evidence is still needed.
Define the question
Start with the decision, prediction or task that matters. Identify the data and what a useful result would look like.
Test the evidence
Check data quality, establish a simple baseline and evaluate limitations before adding complexity.
Agree a pilot
Set a clear scope, named users and acceptance criteria. Agree the handover and support before starting.
Scope, fees, timing and support are agreed in writing before work starts.
Emil Jakobsen
Copenhagen, Denmark
The person behind the work
Delivery experience.
Practical curiosity.
More than eleven years of enterprise delivery at Siemens, ZEISS and TÜV SÜD. Most recently, my work includes a SuccessFactors implementation covering 28,000 employees.
Across those roles, I have enjoyed understanding complex processes, making requirements clearer and turning that understanding into useful changes. I now study machine learning and data science and develop the programming skills to build more of the solution directly.
Start with the task
What could your data
help you understand?
Tell me about the question, the available data and the decision or task you want to improve.
emil@jakobsenanalytics.dk
A few lines about the problem, the data and a useful outcome are enough.
Please leave confidential documents out of the initial enquiry. We can agree a suitable way to share them.