H Company CEO, Gautier Cloix: Beyond the Hype Leading Agentic AI into the Enterprise

Company & Leadership Context

• The interviewee moved from New York to France three weeks prior to join H-Company as CEO.
• H-Company was founded by former DeepMind employees; the interviewee joined after original founders departed to pursue theoretical research.
• The company secured a record-breaking €220 million seed round, described as the largest in Europe at the time.
• The interviewee previously spent 10 years at Palantir, including a deployment for French intelligence following the 2015 Paris attacks.

Core Technology & Strategic Focus

• H-Company builds "computer user agents" trained via human interaction data, enabling them to navigate software without predefined APIs or tooling instructions.
• This approach differs from competitors using "LLMs plus tools" (APIs/MCP); H-Company's agents act natively on browsers, desktops, and mobile apps.
• The company claims to be the global leader on the "Web Voyager" benchmark for computer use.
• H-Company's model, "Holo One," is reportedly developed by a team of 30 top-tier AI researchers and claims performance surpassing US and Chinese counterparts.
• The model architecture prioritizes cost-efficiency, operating at approximately 10x lower cost than standard models, which reduces energy emissions.
• The strategy involves a single generalist model fine-tuned continuously based on deployment feedback (e.g., specific calendar interaction patterns) rather than building siloed agents per software.

Operational Model: Forward-Deployed Engineers

• The CEO is introducing a "forward deployed engineer" (FDE) model to bridge the gap between complex enterprise environments and AI capabilities.
• FDEs are hybrid roles combining business acumen, technical coding skills, and customer proximity to iterate on solutions in situ.
• This model avoids building custom one-off software; instead, insights from customer deployments are fed back to refine the core generalist model.
• Recruitment targets candidates with entrepreneurial mindsets, such as engineers who started companies or dropped out of studies to pursue projects, rather than traditional career paths.
• The FDE culture requires autonomy, rejecting traditional management hierarchies to foster rapid iteration and problem-solving.

Market Strategy & Geography

• Initial customers are concentrated in Europe to ensure rapid iteration with the internal team, specifically targeting large legacy enterprises (e.g., airlines, banks, healthcare).
• The US is a stated priority due to its competitive intensity; signing contracts there averages three weeks compared to 14+ months in France.
• The company explicitly rejects a "ladder" strategy of selling only in one country to avoid stagnation.
• Confirmed interest from investors and potential enterprise customers includes LVMH, Francaise Desjeux, Amazon, Samsung, and UiPath.
• Government sector work is acknowledged as high-impact (e.g., reducing administrative burdens in healthcare) but deprioritized due to long sales cycles unless a crisis occurs.

Talent & Competitive Landscape

• The "talent war" is intense, with competitors like Meta aggressively poaching engineers with high compensation offers.
• H-Company retains top talent through equity ownership and the ambition of building the next trillion-dollar company in computer use.
• The interviewee argues that their natively AI-first culture and FDE model are difficult for established companies to replicate.
• The company plans to onboard 2–3 "big" customers by the end of the year to perfect the product experience before scaling.
• Target markets focus on established companies with legacy, complex IT environments rather than AI-native startups with modern stacks.

Forward-Looking Statements

• The CEO predicts that AI will drastically reduce administrative tasks in healthcare, potentially shifting doctors' administrative time from 70% to near 0%.
• Future growth depends on the ability to maintain low token costs while scaling agents to handle complex, multi-step workflows on mobile and desktop.
• The trajectory involves moving from a small, focused customer base in Year 1 to a broader scale in Year 2.
• The interviewee notes that AI's potential to reduce software development costs does not diminish the need for FDEs, as the primary challenge remains navigating complex, "dirty" enterprise data and policies.