About

A small team that works on hard problems

VectorVirtue was founded in Prague in 2020. We started as a two-person research project and grew into a consulting practice when it became clear that most businesses needed help navigating AI — not more tools added to their stack.

Why we exist

Most AI projects in business fail not because of the technology but because of the gap between what vendors promise and what organisations can actually absorb. The tooling is genuinely good now. The problem is implementation.

We built VectorVirtue to close that gap — by working on the implementation side rather than the product side. We don't sell software. We build and deploy AI solutions using the best available tools for each situation, without any financial interest in which vendor wins.

That independence matters. When we recommend an architecture, it's because it fits the problem — not because we have a reseller agreement.

The founder

Lukas Novotný, Founder and Lead Consultant at VectorVirtue
Lukas Novotný
Founder & Lead Consultant

Lukas spent eight years in applied ML research at a Prague university before moving into industry in 2018. He led AI initiatives at a mid-size Czech manufacturing group before founding VectorVirtue in 2020. His background is in NLP and process modelling, and he leads all client assessments and strategy engagements.

The team

VectorVirtue has a core team of seven people — a mix of ML engineers, integration specialists, and project managers. We also work with a small network of domain experts for sector-specific projects (manufacturing, logistics, professional services).

We keep the team deliberately small. Larger teams mean more coordination overhead and more distance between the people doing the work and the people talking to clients. We'd rather take fewer projects and do them properly.

How we work

Vendor independence

We have no reseller agreements, referral arrangements, or financial relationships with any AI vendor. Our technology recommendations are based on fit, not fees.

Honest scope

We tell clients when AI is not the right answer, or when their data isn't ready for what they want to build. That's not a loss for us — it's the job.

Your team learns

Every implementation includes knowledge transfer. We don't want clients dependent on us to maintain what we build. That dependency is a risk to them, not a feature for us.

Measured outcomes

We agree on metrics before we start and report against them honestly. If the numbers aren't moving, we say so and diagnose why before continuing.