Currently on gardening leave. Interested in working on the hardest technical problems in the machine learning and optimisation space, either as an employee, freelancer, or founder. Also interested in pre-seed investing in projects I feel I can be of use to.
Work in the technology unit on several modernisation projects, both frontend and event-driven backend architectures. Renewed marimekko.com, launched two mobile applications, and a storefront dedicated to personalised fashion. Primary responsible for critical systems including payment processing (charges, holds, automatic refunds) and high-availability hosting.
Primary responsible for country-level technology strategy, with specialisations for machine learning applications. Client work, including bidding algorithms for second-price advertisement auctions under partial information for ten-digit annual budgets.
Consulting work & partner at a machine learning specialist house. Implemented several real-time learning and batch systems for non-stationary problems across telecommunications (churn reduction and treatment) and the fashion industry (real-time webpage optimisation).
Frontend and backend engineering, focusing on cloud systems. Reactive programming for the largest ecommerce business in Finland. Built some of the first real-time contextual bandit learning systems for high volume recommendations in the Nordics. Thrice awarded for performance in annual company ceremonies.
MSc, Computer Science, University of Helsinki; major in software systems, extended minor in mathematics and statistics. Natively trilingual (Swedish, Finnish, English).
Finnish Defence Forces reservist.
Improving the Delivery Cycle: A Multiple-Case Study of the Toolchains in Finnish Software Intensive Enterprises, IST, 2016
The Highways and Country Roads to Continuous Deployment, IEEE Software, 2015
A Behavior Marker tool for measurement of the Non-Technical Skills of Software Professionals: An Empirical Investigation, SEKE, 2015
Examining the Structure of Lean and Agile Values Among Software Developers, XP, 2014
Primary domain: online (incremental) machine learning systems -- algorithms that update model parameters per-observation in O(1) amortised time rather than requiring full batch retraining. Applied focus on contextual bandits, off-policy evaluation (IPS, DR, doubly-robust estimators), and reward estimation under partial feedback. Experience in causal estimation and author of Formative, a causal estimation library. Authored the Vowpal Wabbit off-policy evaluation tutorial in the official documentation.
Secondary domains: linear and mixed-integer optimisation (LP/MIP formulations for scheduling, routing and resource allocation), time-series and signals analysis (spectral methods, online anomaly detection), and distributed backend architectures for real-time inference (streaming ingestion, feature stores, sub-millisecond serving layers).
Public talks on contextual bandits, bandit algorithms, and cloud-native ML at Aalto University, Junction Helsinki and Helsinki Reinforcement Learning meetup.