Until now, corporate training has been structured around a relatively stable model composed of a centralized platform, standardized content, and courses consumed at regular intervals. This model, embodied by LMS and LXP, has made it possible to industrialize training.
The gap between the investment made and the real impact is now difficult to ignore. Engagement rates remain low, knowledge retention limited, and training often remains disconnected from operational situations. In other words, the current learning stack fulfills an administrative function more than a strategic role.
It is this gap that some startups are today trying to address, by relying on a broader shift, that of software towards workflow.
From platform to flow: a paradigm shift
The first shift is that of distribution. Historically, training was based on platform logic: the user had to connect, navigate, then consume content. This model assumes prior intention and relies on individual discipline that is difficult to maintain in fragmented work environments.
🚨 SMARTJOBS
- MISTRAL – Account Executive, Enterprise, France – Paris
- ANTHROPIC – Startup Partnerships – France & Southern Europe
- CONTEXT – HR Director – Human Resources Director
- ECOLE POLYTECHNIQUE – Director/Deputy Director of International Relations (F/M)
- CLAROTY — Sales Development Representative
- FRACTTAL — Account Manager (France)
- BRICKSAI — Founding Growth Manager
👉 Find all our offers on the DECODE MEDIA Jobboard
📩 Are you recruiting and want to strengthen your employer brand? Discover our partner offers
A new approach is emerging, consisting of integrating training directly into the tools used on a daily basis, whether messaging, collaborative tools, or business environments. This change in integration modifies the very nature of training.
It is no longer a question of accessing content, but of receiving contextualized information, when it becomes useful.
The end of the catalog as central unit
The second shift concerns content. LMS were designed like catalogs: an accumulation of modules, structured by theme, accessible on demand. This logic remains relevant for organizing information, but it does not guarantee engagement or learning.
The rise of artificial intelligence introduces a new logic, that of large-scale individualization. Content is no longer just stored and distributed, but activated according to the context, role, and level of the user.
In this model, the value no longer lies in the quantity of content available, but in the ability to deliver the right information at the right time.
From a training logic to a performance logic
This technical shift is accompanied by a more strategic shift.
In this dynamic, training can be seen as a lever for operational performance, especially since it ceases to be a moment distinct from work to become a continuous process, integrated into the activity.
Of course, this development forces human resources departments to reconsider their role. It is no longer just a question of structuring a training offer, but of designing a system capable of constantly supporting the development of skills.
In this context, stack learning tends to become an infrastructure, connected to the other building blocks of the company
A stack to rebuild, not optimize
Faced with these developments, one temptation is to add technological layers to existing systems. This incremental approach quickly reaches its limits.
The subject is not to improve the existing, but to rethink architecture.
Three structuring questions emerge:
- Where is the training located: in a dedicated platform, or in everyday tools?
- How is it triggered: at the user’s initiative, or contextually?
- On what is it measured: on completion rates, or on performance indicators?
These trade-offs redefine the learning function itself, which comes closer to real use.
AI as a catalyst, more than a solution
Artificial intelligence plays a central role in this transformation, but its contribution is often misinterpreted. It is not limited to accelerating the production of content, but above all allows it to contextualize, orchestrate and activate learning. In other words, it affects distribution more than the content itself.
This point is decisive, because it explains why historical players, designed around a catalog logic, struggle to fully integrate these new approaches. Their initial architecture does not correspond to this real-time activation logic.
An ongoing reorganization of the HR Tech market
The corporate training market is currently experiencing a phase of accelerated restructuring, both in terms of its scale and the nature of the transformations underway. On a global scale, e-learning already represents a market estimated at around $320 billion in 2025, with annual growth close to 14%, while the more specific LMS segment is expected to exceed $100 billion by 2034. Behind these volumes, the competitive structure is evolving rapidly. On the one hand, the historical players, Cornerstone, Docebo, 360Learning, are trying to integrate artificial intelligence building blocks without calling into question their architecture centered on the catalog. On the other hand, a new generation of players, such as Sana Labs, eduMe or Spekit, is striving to move training towards the workflow, by integrating 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 the learning and development market appears to be one of the most natural use cases of artificial intelligence. Between the consolidation of existing platforms and the rapid arrival of new entrants, the sector is fragmenting while redefining itself around a central question: should training remain a software product, or become an invisible infrastructure for performance at work.
It is precisely in this reconfiguration that the emergence of new players falls, and the raising of 1.8 million euros carried out by Blify illustrates this more structural dynamic.
Founded in 2025 and based in Boulogne-Billancourt, the company is developing an approach consisting of integrating training directly into daily work tools, relying on a multi-agent AI infrastructure. It offers what it describes as a “Learning Operating System”, aiming to deliver contextualized content throughout the activity.
Blify was founded by three profiles from HR Tech and SaaS: Clément Lhommeau (formerly 360Learning), Tristan Vié (ex-JobTeaser) and Minh-Tu Hua (ex-Alan). His round table is led by AFI VenturesVentech seed fund, alongside notably Kima Ventures.
