Why HR leaders must fully rethink their learning stack (and what Blify’s €1.8 million funding round reveals)

Until now, corporate training has been structured around a relatively stable model built on a centralized platform, standardized content, and learning paths consumed at regular intervals. This model, embodied by LMS and LXP systems, has enabled the industrialization of training.

The gap between investment and actual impact is now increasingly difficult to ignore. Engagement rates remain low, knowledge retention is limited, and training often stays disconnected from operational realities. In other words, the current learning stack fulfills an administrative function more than a strategic role.

It is this disconnect that a new generation of startups is attempting to address, leveraging a broader shift—from software as a destination to software embedded in the workflow.

From platform to flow: a paradigm shift

The first shift concerns distribution. Historically, training has relied on a platform logic: users had to log in, navigate, and then consume content. This model assumes prior intent and depends on a level of individual discipline that is difficult to sustain in fragmented work environments.

A new approach is emerging, integrating learning directly into the tools used daily—whether messaging platforms, collaborative tools, or business applications. This shift in integration fundamentally changes the nature of training.

It is no longer about accessing content, but about receiving contextualized information at the exact moment it becomes useful.

The end of the catalog as the central unit

The second shift concerns content. LMS platforms were designed as catalogs: accumulations of modules, structured by topic and accessible on demand. While this logic remains useful for organizing information, it does not guarantee engagement or learning outcomes.

The rise of artificial intelligence introduces a new paradigm: large-scale individualization. Content is no longer simply stored and delivered, but activated based on context, role, and user proficiency.

In this model, value no longer lies in the quantity of available content, but in the ability to deliver the right information at the right time.

From training logic to performance logic

This technical shift is accompanied by a more strategic repositioning.

Training can increasingly be seen as a lever for operational performance, as it ceases to be a distinct moment separate from work and instead becomes a continuous process embedded within daily activity.

Naturally, this evolution forces HR leaders to reconsider their role. It is no longer just about structuring a training offering, but about designing a system capable of continuously supporting skills development.

In this context, the learning stack tends to become an infrastructure, interconnected with other enterprise systems.

A stack to rebuild, not optimize

Faced with these changes, there is a temptation to simply add new technological layers onto existing systems. This incremental approach quickly reaches its limits.

The challenge is not to improve what already exists, but to rethink the architecture entirely.

Three structuring questions emerge:

• Where does learning take place: within a dedicated platform, or within everyday tools?
• How is it triggered: by user initiative, or contextually?
• How is it measured: through completion rates, or performance indicators?

These trade-offs redefine the learning function itself, bringing it closer to real usage.

AI as a catalyst, not a solution

Artificial intelligence plays a central role in this transformation, but its contribution is often misunderstood. It does not merely accelerate content production; it enables the contextualization, orchestration, and activation of learning. In other words, it impacts distribution more than content itself.

This distinction is critical, as it explains why legacy players—built around a catalog logic—struggle to fully integrate these new approaches. Their original architecture is not designed for real-time activation.

A market in the midst of recomposition

The corporate training market is currently undergoing a phase of accelerated recomposition, both in scale and in the nature of the transformations underway. Globally, e-learning already represents an estimated $320 billion market in 2025, growing at nearly 14% annually, while the more specific LMS segment is expected to exceed $100 billion by 2034.

Behind these figures, the competitive landscape is shifting rapidly. On one side, incumbent players such as Cornerstone, Docebo, and 360Learning are attempting to integrate AI capabilities without fundamentally rethinking their catalog-based architecture. On the other, a new generation of companies—such as Sana Labs, eduMe, or Spekit—is moving learning into the workflow by embedding it directly into operational tools.

This trend is reinforced by the emergence of AI-native solutions capable of generating, personalizing, and activating content in real time, in a context where learning and development appears to be one of the most natural use cases for artificial intelligence. Between the consolidation of existing platforms and the rapid emergence of new entrants, the sector is fragmenting while being redefined around a central question: should training remain a software product, or become an invisible infrastructure for performance at work?

Blify: a signal of structural change

It is precisely within this reconfiguration that new players are emerging, and Blify’s €1.8 million funding round illustrates this broader structural dynamic.

Founded in 2025 and based in Boulogne-Billancourt, the company is developing an approach that integrates learning directly into everyday work tools, relying on a multi-agent AI infrastructure. It offers what it describes as a “Learning Operating System,” designed to deliver contextualized content continuously throughout daily activity.

Blify was founded by three profiles from HR Tech and SaaS: Clément Lhommeau (formerly at 360Learning), Tristan Vié (formerly at JobTeaser), and Minh-Tu Hua (formerly at Alan). The round was led by AFI Ventures, Ventech’s seed fund, notably alongside Kima Ventures.