Mendo

What is it?

It began in 2021 when Quentin and Alexandre saw the challenges enterprises faced in adopting generative AI. They noticed that existing approaches – one-off training, static courses, generic rollouts – led to low adoption and poor ROI on AI investments. Mendo was created as an adoption layer that embeds directly into the GenAI tools teams already use – Copilot, ChatGPT, Gemini, Claude – guiding usage, surfacing relevant use cases by role and giving leadership clear visibility into what’s working.

Why we invested

The founders

Quentin moved to Madagascar at 13, where he saw first-hand how access to the right skills and knowledge can reshape a person’s opportunities to fulfil their potential. With a passion for entrepreneurship, he studied data and finance at McGill University and earned a Master’s in management from ESSEC. After a brief stint in investment banking, he co-founded Mendo.

Alexandre, the first in his family to attend university, grew up in a challenging area of Paris. He saw early on that technology could be the great equaliser, giving people access to capability that their circumstances alone would not. Alexandre studied physics, management and data science, and worked as a Technical Project Manager in New York City before starting Mendo.

The problem

The future of work is being fundamentally reshaped. Agents are moving from experimentation into production, job roles are being redefined in real time, and the pace of change is outstripping any traditional training approach. Organisations face not just an adoption gap but a continuous transformation challenge: how do you equip every employee to work alongside AI systems that are themselves evolving week by week? Without a structured adoption layer, enterprises are left with scattered usage, invisible ROI and a workforce falling behind faster than they can course-correct.

The solution

Mendo has built a GenAI adoption layer that lives inside the AI tools teams already use, rather than introducing yet another platform. It personalises learning and use case discovery by role and context, embeds guidance directly in the flow of work, and provides organisations with real data on adoption, usage patterns and ROI. As enterprises move toward agentic workflows, Mendo also supports each agent deployment end-to-end, from scoping and onboarding to structured rollout and feedback loops, ensuring human-agent collaboration becomes a lasting organisational capability, not a one-off pilot. Currently compatible with Microsoft Copilot, ChatGPT, Gemini, Claude and custom internal AI environments, Mendo serves 100+ enterprise clients including EY, PwC, Novo Nordisk and Crédit Agricole.

The opportunity

Enterprises are pouring money into GenAI tools, but adoption is stalling. Microsoft 365 has over 400 million commercial users, yet only around 3% have converted to paid Copilot seats. Across the board, Deloitte’s 2026 State of AI report found that 60% of employees now have access to AI tools, but fewer than 60% of those with access regularly use them, and only 20% of organisations say their workforce is highly prepared. OpenAI’s own enterprise data shows a 6x productivity gap between top AI users and the median employee. What’s missing is an adoption layer that embeds directly into the AI products teams already use, turns scattered experimentation into structured capability, and gives leadership real visibility into what’s working and what isn’t. That layer doesn’t exist in the AI products themselves, and traditional training can’t keep pace. This is the category Mendo is building.