Replenit Nabs USD 2.5M to Reinvent Retail Decision-Making With AI

• Warsaw-based Replenit attracted USD 2.5M of Pre-Seed funding to build AI-driven retail decision-making infrastructure.
• The startup was founded by former InsiderOne executives, experienced in scaling global martech operations internationally
• The solution helps retailers make individualized real-time decisions using behavioral reasoning-based AI
• The new funding enables Replenit’s US expansion, deeper AI research, and broader enterprise retail adoption

This April, Replenit—the Warsaw-based AI decision-making enabler for retailers—raised its USD 2.5M Pre-Seed round of investment. One of Poland’s most active VC firms Movens Capital (invested in Talkie AI, among others) co-led the round, joined by vastpoint. Other participants included Logo Ventures, Digital Ocean Ventures (invested in Demoboost, among others), Finberg, Caucasus Ventures (invested in KLIPY, among others), and Mati Staniszewski of ElevenLabs.

From Martech Unicorn Experience to a New AI Venture

Replenit was founded in 2025 by a team of six Turkish entrepreneurs with more than 40 years of combined experience scaling B2B SaaS and martech companies internationally, including to unicorn level. They had previously helped scale the Turkish martech company Insider (presently InsiderOne) to USD 2B across 26 countries.

Working with hundreds of enterprise retailers in Europe, the UK, the US, MENA, and Latin America, Ilyas Kurklu (CEO) and his team repeatedly encountered the same problem: Companies had data and execution tools in spades, but no real decision intelligence.

The Retail Industry’s Missing Intelligence Layer

‘Retailers are sitting on tons of customer data, but very few can turn it into real-time decisions that drive revenue. Replenit is building the AI layer that closes that gap, helping brands act at the exact moment a customer is ready to buy again. With a team that has already scaled global martech platforms and early proof with major retailers, we believe Replenit has the potential to define a new category in retail infrastructure,’ Movens Capital’s investment director Łukasz Lewandowski firmly believes.

‘This problem was never theoretical to us. We saw the technical and operational realities of enterprise retail firsthand, and intuitively understood retailers’ need for a new intelligence layer rather than another tool,’ Mr Kurklu tells ITKeyMedia.

According to him, retail technology has evolved in layers. Data systems that collect customer signals came first, and were followed by orchestration tools like CRM and marketing automation that execute actions. After this, retail-specific innovation stalled. The industry started operating under the assumption that prediction models and segmentation would remain sufficient to drive decisions, and most retailers still rely on static rules, broad segments, and campaign calendars to manage highly dynamic customer behaviour, which is largely based on historical patterns and fails to capture individual context or intent.

The Replenit team sees most recommendation systems or next-best-action tools as fundamentally reactive, relying on some combination of predefined logic, historical patterns, and limited optimization frameworks. While retailers already have tools to send emails, push notifications, and run campaigns, their problem is deciding what to send, when, and to whom.

This gap leads to:

• more than 50% of consumers saying personalisation feels off-target,
• although 90% of marketers consider first-party data critical, fewer than 30% can integrate it effectively across channels
• meanwhile, AI search is already reshaping how shoppers find and evaluate products, on track to influence USD 750B in consumer spending by 2028

Beyond Prediction and Toward Real-Time Decision-Making 

Now that AI has matured significantly, the Replenit team spotted the opportunity to move beyond prediction and start understanding behaviour in a more contextual, human-like way, and to generate decisions at the level of each individual customer.

‘The shift is equal parts technological and conceptual. Retail is moving from rule-based execution toward more intelligent, decision-driven systems,’ Mr Kurklu states.

A Theory-of-Mind Approach to Retail AI

In this situation, Replenit introduces a reasoning layer built on a theory of mind approach. Instead of asking what worked before or what product is similar, its technology seeks to understand why a behaviour happened, what it signals about intent, and what it means in context—bordering how a human would interpret it. Operating at the level of each individual customer-product relationship, the system continuously determines what should happen next. This goes beyond suggesting options, making and triggering decisions through existing systems.

‘At the core of our approach is a theory-of-mind-inspired framework. Rather than focusing purely on customer actions, we understand the motive behind those actions and what they signal about intent and context. This reasoning layer generates synthetic data that helps enrich both product and customer relationships. This allows it to move beyond surface-level correlations and reason about what the behaviour actually means,’ Mr Kurklu explains.

An AI Decision-Maker Above the Existing Stack 

‘We are not an optimization layer within a CRM. We sit above the stack as the intelligence layer that drives decisions across all channels and tools. Replenit acts like an AI decision-maker sitting on top of existing tools. It looks at customer behaviour, understands what that behaviour means, and decides the best next action. It considers factors like when someone is likely to buy again, what product they might need, or when they might be at risk of dropping off. It then triggers those actions through the retailer’s existing systems, so retailers do not need to retrofit or replace anything. Everything becomes more precise and personalized,’ the founder continues.

From Movens Capital’s perspective, retail is still largely operating on outdated paradigms — static rules and backward-looking analytics — while the real challenge lies in making high-quality decisions in real time. Replenit, in turn, addresses this gap by building an AI-driven decision layer that focuses not just on predictions, but on executing the right action, at the right moment, for each individual customer.

The fund’s partner Artur Banach lists the factors that stood out to his team:

• A clear shift from prediction to decision-making – a layer that is still largely missing in modern retail stacks
• Strong early traction, with over 30 global brands already onboarded
• Tangible impact on revenue, including meaningful improvements in upsell performance
• A founding team with proven experience in building and scaling global martech platforms
• Day-one global ambition, particularly with a focus on the US market

‘Importantly, Replenit is not just optimizing campaigns in isolation, but fundamentally improving the quality of each commercial decision. While that may sound subtle, in practice it represents a meaningful shift in how retail organizations leverage data,’ Mr Banach remarks.

The existing infrastructure with which Replenit integrates includes Bloomreach, Braze, Databricks, Klaviyo, and Salesforce. Integrating across multiple martech stacks, the hardest bottleneck has been that, while retailers have a lot of data, it is often not clean and is fragmented across different systems and operations. To tackle this, Replenit built in a data enrichment layer to work even with limited or imperfect data and still generate meaningful decisions.

‘We fill the gap in the data through our reasoning layer, which helps enrich existing data. When behavioural data is limited, the system generates additional context to strengthen decision-making. This enrichment is not only used within Replenit’s own operations, but is also fed back into the retailer’s data warehouse, improving the overall quality and usability of their data over time,’ Mr Kurklu comments.

Replenit already supports a range of real-time, individual-level decisions across the customer lifecycle:

• replenishment and repeat-purchase decisioning
• cross-sell and upsell
• promotion and individual level churn prediction
• substitute product recommendation
• and AI-powered product data enrichment and decision support.

Real-World Revenue Impact

These are already deployed with enterprise retailers and delivering measurable impact, like significant increases in post-purchase revenue. The startup reports that L’Occitane en Provence recorded a 235% increase in post-purchase revenue after deploying Replenit’s engine. Meanwhile, iBOOD — one of Europe’s best-known flash-deal retailers — now attributes 6.3% of total company revenue to Replenit-driven decisions. 

To emphasize its confidence, Replenish includes an explicit 10x ROI guarantee in its contracts, with a contractual exit clause if results are not delivered, and cites that no customer has invoked it to date.

Scaling AI Research, Product Development, and US Expansion

According to vastpoint’s founding partner Karolina Kukielka, her team backed Replenit mostly because of a very compelling opportunity at the intersection of e-commerce growth and AI-driven retention. 

‘The team also demonstrated a strong grasp of the problem space and early traction that validated their core value proposition. For vastpoint, Replenit represents the kind of pre-seed bet we’re drawn to: a clear pain point, a scalable SaaS model, and a founder team with the right mix of domain expertise and execution energy to grow across the European e-commerce market,’ Ms Kukielka shares.

Mr Kurklu further specifies that Replenit’s new funding is meant to strengthen three core areas of the business over the next phase:

• A major priority is deepening product development and AI research, particularly to expand the capabilities of the decision engine and reinforce its scientific foundation as the startup is growing our engineering and research teams in Poland and the Netherlands to support that development and scale the platform more effectively;
• The company is also preparing for a more structured expansion into the US, including building a local team and establishing a stronger commercial presence there throughout 2026;
• Beyond 2026, wthe team targeting 8-digit ARR, reaching more than 100 global enterprise clients where Replenit is selected as their intelligence layer, and expanding our coverage beyond customer engagement to make our decisions accessible to each department of the retailers.

‘We are implementing a scientific approach, working closely with academics in consumer behavior and psychology so that Replenit is not only a technology layer, but also built on a strong scientific foundation. Our overarching ambition is to become the central intelligence layer for retail, covering all commercial decisions across the customer lifecycle,’ Mr Kurklu concludes.

The emergence of Replenit and its further growth reflect a shift in retail technology, where competitive advantage increasingly relies on making intelligent real-time decisions from the collected customer data. Positioning itself as an AI reasoning layer rather than another marketing tool, the Warsaw-based startup is tapping into one of the industry’s biggest unresolved challenges: turning fragmented data into measurable commercial outcomes at scale. With fresh backing, Replenit has the momentum to expand its research, accelerate international growth, and potentially define a new category in AI-powered retail infrastructure.