Businesses
Recurring reports, deeper analysis or a specialised model for your own data, from €100 a month depending on the data and the complexity.
See the packages →Data analyst and data scientist · Riga, Latvia
Girts Liepins
I turn messy business and market data into reports, forecasts and models that people actually use, in banking, energy and retail. NordShifts, this site, is where my own electricity and equity research is published.
- 8+ years in data roles
- Banking, energy and retail
- Python, SQL, Power BI, machine learning
- Latvian and English
The vision
Decisions on energy and markets should rest on numbers anyone can check
NordShifts started from a simple question: how good are the forecasts people rely on for electricity prices and energy shares, really? Most tools show a confident line and no record of how often it was right.
This platform does the opposite. Every forecast is frozen before the outcome is known, scored every day against a simple no-change benchmark, and published, including the cases where the model does not beat it. The aim is a Baltic and Nordic research desk that households, companies and investors can trust because they can check it.
Who it is built for
Households and small businesses
See tomorrow's electricity price in their own zone and language, and the cheapest hours to use power.
Energy companies and suppliers
Forecasts, demand and risk figures through an API, with an honest track record behind every number.
Investors and analysts
Battery, PPA and security-of-supply screens, and research on about 100 listed clean-energy companies.
Banks and lenders
Collections and lending analytics from someone who has built them inside financial groups.
Who should hire me
- Companies with data they don't use yet, that want a monthly report, a forecast or a model they can rely on.
- Energy, finance and retail teams that need analysis tested properly before decisions are made on it.
- Employers looking for a senior data analyst or data scientist who builds complete solutions, from SQL and Power BI to machine learning in production.
- Investors and partners who want to take the NordShifts platform further.
How I can help
Investors and partners
NordShifts publishes Baltic and Nordic electricity forecasts, battery and PPA screens and clean-energy company research, each with its measured track record.
Open the investor desk →Employers and recruiters
Senior analyst experience in banking and collections, Power BI reporting built from scratch, and production forecasting models in Python.
Download the CV →What I do
Reporting and BI
Power BI and Tableau reporting systems, data models and SQL that stay fast and match the source systems.
Forecasting and machine learning
Price, demand and sales forecasts, from regression to gradient boosting and deep learning, always tested against a simple benchmark.
Financial and market analysis
Collections and lending performance, company financials, equity screens and electricity market research.
Automation and data pipelines
Python scripts, APIs, web scraping, VBA and Power Query that replace manual work.
Own projects
Two products I designed, built and run myself, from the data and models to the servers.
NordShifts
Visit nordshifts.comElectricity market and energy-company research platform for the Baltic and Nordic region.
What I built
- Day-ahead price forecasts for every Baltic and Nordic zone, with a public daily track record
- Battery, PPA, carbon-intensity and security-of-supply screens for investors
- Profiles of about 100 clean-energy companies with forecasts, backtests and quarterly ratios
- A data API with keys and usage limits, in eight languages, on PostgreSQL and Docker
LuminShop
Visit luminshop.euAn online fragrance shop with its own pages for each language.
What I built
- The shop itself, with indexable product pages per language
- A shopping assistant and a fragrance matcher that suggest scents
- A research ranker and a merchant product feed
- An SEO plan for Latvian, Lithuanian, Estonian, Finnish and Polish
Experience
The short version. The CV has the details.
2024 – now
Senior Data Analyst, Strategy & Analytics · Pactum Collections
Collections performance reporting by country and agency, reconciled with agency data, on dedicated tables built with the data engineers.
2022 – 2024
Power BI Developer and Data Analyst · Sefinance
Built the company's Power BI reporting for marketing, sales and finance from scratch, with Python integrations for Google Analytics and Ads.
2022
Power BI Owner and Data Analyst · Madara Cosmetics
Purchasing and manufacturing automations from MSSQL with Power Query and VBA.
2020 – 2022
Business Analyst and RPA Developer, Process Innovation · Samsung Electronics Baltics
Python price collection from Baltic e-stores, sales forecasting, the SAP HANA migration and a team of three.
2018 – 2020
Business Information Specialist · SEB Global Services
Industry research for rating reports and the KYC teams' first Tableau and Power BI reporting.
2018
Electronic Data Specialist · AC Nielsen Baltics
Retailer data feeds and the SQL that loaded them.
Bachelor in Business Management, Ventspils University College, 2018
Tools and skills
- Python
- SQL
- Power BI
- Tableau
- Machine learning
- Deep learning
- AI engineering
- Power Query
- Excel and VBA
- SAP
- JavaScript and React
- Node.js
- Financial data and statements analysis
- Market research
- Data engineering
- Product ownership
- Leadership
- Data-driven M&A advice
About NordShifts
NordShifts publishes day-ahead electricity prices and forecasts for the Baltic and Nordic zones, research on European and US energy companies, and a data API. Every model is compared with a simple no-change benchmark, and the results are public, including where the models don't beat it.
More about the platform
What this is
NordShifts is a day-ahead price advisor for households and small businesses in the Baltic and Nordic bidding zones. The first markets are Latvia, Lithuania and Estonia, because that is where a clear public tool in the local languages is still thin.
The same auction covers Finland, Sweden, Norway and Denmark, so those languages and zones are here too, each with its own pages.
What this is not
It is not an electricity supplier, not a bill forecast, and not a promise of a percentage saving. The savings figure is the energy-price gap on one delivery day.
It is not an automated trading desk. Battery advice on the page is a transparent daily cycle, stated with its assumptions.
Why these languages
People search for elektrības cenas, elektros kainos, elektri hinnad, sähkön hinta, elpriser and strømpriser, not for an English product name. Each language has its own URLs, titles and market notes.
A translation agent using the OpenAI or Claude API can draft the next note or a new language pass. You approve it before it replaces a live page. Rough unreviewed copy is a good way to lose a market where readers notice.
What stays public
The auction, the method and the persistence baseline are on the site. A richer forecast earns a place by beating that baseline on a history we keep.
Small numbers, shown plainly, are the point. The accuracy page is allowed to look ordinary.
Other work
NordShifts is my main product. Before and beside it I have built and still run other projects; they are not part of this site, but I am happy to talk about them.
Equity and clean-energy research
Financial statements, ranking models and walk-forward backtests for European and US listed companies, and a portfolio lab.
Freelance data analysis
Reporting, forecasting and machine-learning work for banking, collections, retail and energy teams.
LuminShop
A fragrance shop with its own shopping assistant, scent matcher and product feeds.
Football data platform
Player-potential models and comparison charts for smaller European leagues.
Have data and a question?
Tell me what you want to find out. I reply within two working days.