Why the marketing stack is being rebuilt from scratch
Our investment conviction, how we evaluate founders, and what we do after we write the first check.
Generative AI is not augmenting the marketing stack. It is restructuring it. The assumptions that underpinned every layer of the stack — from content production timelines to audience segmentation to creative testing cycles — are being invalidated faster than incumbent platforms can adapt. This creates a seed-stage investment opportunity of a scale we have not seen since the SaaS transition disrupted on-premise enterprise software in the early 2010s.
The opportunity is at the infrastructure layer, not the application layer. When a platform technology shifts this fundamentally, the infrastructure must be rebuilt before the application layer consolidates. We back startups that are building the infrastructure: content generation pipelines, AI-native personalization engines, creative intelligence and attribution systems. We invest at seed stage, with initial check sizes of €1.5M–€3M, because that is where the infrastructure decisions are made — before category leaders emerge and valuations price in the outcome.
The European market creates a structural advantage for this firm. European buyers have materially different procurement structures, compliance requirements, and enterprise sales cycles than their US counterparts. Founders who understand how to sell into European enterprise — specifically GDPR-aware data handling, longer contract cycles, procurement committee dynamics in France, Germany, and Iberia — have a distribution moat that US-headquartered competitors consistently underestimate. We back founders who are building with European market fluency from day one, not retrofitting it after US product-market fit.
Within AI-powered marketing technology and content intelligence, we prioritize four categories: content generation at machine scale (replacing the content operations function, not just accelerating it), AI-native audience personalization (not rules-based segmentation with AI labels), generative creative and attribution intelligence (closing the loop between creative output and measurable business outcomes), and multilingual and cross-market content infrastructure (a specifically European opportunity). These categories share a structural property: the founding team's model quality creates a compounding data advantage that becomes harder to replicate as the product scales. That data flywheel is what we are looking for in a sector: AI for MarTech.
We run concentrated portfolios. Fund I deployed into 8 companies; Fund II is targeting 12–14 investments. Concentration is a deliberate choice — it means we can run structured GTM reviews, join a CTO's architecture decision, or help a founder prepare a fundraise narrative, rather than spreading thin across 40 companies with quarterly check-in calls.
Initial check sizes range from €1.5M to €3M. We reserve 40–50% of each fund for follow-on into our strongest performers. We do not co-invest passively — when we are in a round, we arrive with a specific operating contribution: Tiago's ad-tech distribution network, Inês's European enterprise buyer relationships, or Rui's technical model evaluation work. That contribution is defined before the check is signed.
How We Work with Founders
Go-to-market in European enterprise
Inês has run marketing from employee #15 through IPO at a Lisbon SaaS unicorn and CMO at a Barcelona-based analytics company. Her network of procurement decision-makers, channel partners, and enterprise buyers across Iberia, France, Germany, and the UK is available to portfolio companies navigating their first enterprise GTM motion. We run structured GTM reviews at the 6-month mark with each portfolio company — mapping ICP, refining positioning, identifying the specific buyer personas who close first. This is not generic advice. This is specific introductions to named buyers in named organisations, combined with a methodology for how to structure the early sales motion in European enterprise contexts where the buying committee typically includes legal and IT alongside business stakeholders.
AI model evaluation and content pipeline architecture
Rui spent three years as a founding engineer at a content intelligence platform before completing his PhD in computational linguistics. He runs technical diligence for every new investment and offers ongoing advisory to portfolio CTOs on model evaluation frameworks, NLP pipeline architecture, fine-tuning strategy, and the infrastructure decisions that affect long-term product quality. We hold quarterly architecture reviews for portfolio companies navigating infrastructure scaling decisions. When a portfolio company is evaluating whether to fine-tune a foundation model, build proprietary training pipelines, or license a third-party API, Rui works through that decision with real cost and performance data — not vendor marketing materials.
Fundraising for your next round
Tiago has raised two institutional funds from LPs including European family offices, corporate strategics in the media and advertising sector, and a fund-of-funds. When portfolio companies are ready to raise their Series A, we provide active warm introductions to tier-1 European and US-based growth investors. We do not make introductions until a founder signals readiness — we believe the preparation for a raise is as important as the raise itself. We work with founders on narrative construction, financial model review, and investor-specific positioning in the 6–8 weeks before a raise process opens. Our LP base includes strategics who occasionally lead growth rounds in portfolio-adjacent companies, creating additional optionality for portfolio companies with the right profile.
Technical talent and executive recruiting
Our network of NLP engineers, ML researchers, and product leaders spans the European university and startup ecosystem. We actively assist portfolio companies hiring their first ML engineer or VP of Product. We make direct introductions and assist with compensation benchmarking for early technical hires in the Lisbon, Barcelona, London, and Berlin markets. The AI talent market in Europe is tight, particularly for candidates who combine research depth with production engineering experience. Through Rui's academic network and the broader Telhaverde portfolio ecosystem, we have been able to source candidates who do not appear on LinkedIn job boards — founding engineers and researchers who are open to the right opportunity but are not actively looking.
What We Look For
Distribution as a moat
We prioritize founders who have a specific, defensible theory of how their product reaches and retains marketing and creative buyers — not just a product. Channel strategy is diligenced as rigorously as technical architecture at seed stage.
Technical depth in the core model
We look for founding teams where the NLP or vision model is the product, not just a wrapper. The model must create a compounding data advantage over time. Fine-tuning on customer data that improves output quality with scale is a prerequisite signal for us, not a bonus.
European market fluency
We back founders who understand how the European enterprise buying cycle differs from the US — longer sales cycles, different procurement structures, local data compliance considerations. Founders who have navigated this have a structural distribution advantage that takes US-market-first competitors years to replicate.