How We Invest
Seed and Pre-Seed investing in AI-native logistics and supply chain startups across Europe. $1M–$3M at seed. $500K–$1.5M at pre-seed.
The Thesis
European logistics and supply chain infrastructure runs on software written before machine learning existed at commercial scale. The systems that coordinate freight, manage warehouse operations, predict demand, and route last-mile delivery were built for reliability, not intelligence. They work. They just can't learn.
That window is closing. A generation of founders who grew up inside these systems — engineers who built the forecasting models, operators who managed the warehouses, analysts who ran the procurement cycles — are now building AI-native replacements. Not integrations. Not dashboards. Companies where the machine learning model is the core product and the data flywheel is the business model.
We invest in that category because we've worked inside it. Erik spent seven years building optimization systems at a Nordic 3PL before moving to VP Engineering at a freight-tech exit. Maja ran distribution across 14 European markets before becoming COO at a warehouse automation startup. Anders built demand forecasting models at a grocery retailer and then an AI logistics startup in Berlin before joining Kvistlund. We don't pattern-match against the US playbook. We diligence from experience.
Europe is underinvested in this category relative to its structural advantage. Nordic and DACH supply chains are among the most complex and data-rich in the world. The founders building here have real domain expertise and real access to enterprise buyers. The gap between what's being built and what's being funded is where we operate.
Check size: $1M–$3M at seed. $500K–$1.5M at pre-seed.
$82M under management across Fund I ($35M, fully deployed) and Fund II ($47M, actively deploying).
How We Work with Founders
Maja spent nine years as an operator before joining Kvistlund — managing distribution contracts across 14 European markets at a Scandinavian FMCG group, then as COO at a warehouse automation startup navigating exactly the enterprise sales cycle our portfolio companies face. Her procurement and commercial relationships are available to portfolio founders working their first large-operator deal. We make direct introductions to heads of logistics operations, procurement leads at 3PLs, and supply chain IT buyers at grocery and retail groups. We do this because cold outreach into these organisations rarely reaches a decision-maker — the right introduction saves months.
Anders leads post-investment architecture reviews. As an ML engineer who built demand forecasting at scale, he works directly with founding CTOs on infrastructure decisions — data pipelines, model retraining cadences, cloud cost architecture. We hold standing monthly architecture sessions with technical co-founders. The questions that come up — how to handle concept drift in a live model, how to structure ground-truth labeling at logistics operators, when to move from batch to streaming inference — these are not hypothetical. Anders has worked through them in production environments.
We prepare founders for Series A conversations starting at month 6 post-investment. This includes investor narrative work, financial model review, and warm introductions to relevant Series A investors in Europe and the US. We have ongoing relationships with Tier 1 European growth funds and a small number of US logistics-focused funds. We don't do generic intros — we go to market with a founder when we believe the timing and story are right, and we prepare them for the specific questions those investors ask.
Erik takes a board seat at every seed investment. Board participation means quarterly formal reviews — prepared with an agenda, structured around metrics and milestones — plus async availability for strategic decisions. We are accessible by design, not by appointment. If a portfolio company is navigating a difficult customer negotiation, a key hire decision, or an unexpected technical setback at 11pm on a Tuesday, we're reachable. Small firms with short chains of communication are an advantage here that larger funds can't replicate.
What We Look For
We don't invest in logistics companies that added machine learning to an existing product. We invest in companies where removing the ML model would mean there is no product. The data flywheel has to be the business model — each new customer makes the system smarter in ways that compound over time. This distinction determines whether you're building a defensible AI company or an enterprise software company with a machine learning module.
At least one co-founder must be operating inside Europe at the time of investment. Not "planning to expand to Europe." Not "based in New York with European customers." Building a company for European logistics operators requires understanding European supply chain complexity — multi-modal freight across different regulatory environments, languages, and operator relationships. That understanding is earned by being here.
The best logistics AI companies we've seen are built by people who worked the problem before they wrote a line of code. That doesn't mean every founder needs a decade of operations experience — but it means someone on the team has to have been inside the system they're automating. If a founding team can't explain exactly why their target buyer's current process is painful, from the perspective of someone who has done that job, we won't invest.