Conference Presentation, Fireside Chat
Building AI That Gets Work Done | Gautier Cloix, H Company | RAISE Summit 2026
- AI discourse is projected to shift within one year from technology discussion to practical business impacts and human strategic shifts once the "execution layer" is removed.
- Early adopters of computer use will see repetitive tasks automated, leading to the pursuit of complex, high-value ideas as implementation barriers vanish.
- The company's Annual Recurring Revenue (ARR) is expected to reach a significantly higher level by the end of next year, though no specific figure is provided.
- Cost savings and increased customer volume are anticipated as CFOs benefit from reduced costs and sales teams reduce time spent on CRM and email management.
- European success in the AI sector is predicted through focusing on specific niches that become mainstream rather than competing with high-cost frontier models.
- Model performance is expected to improve by 20% to 30% within three days of any new Nvidia Nemotron model release, contrasting with standard year-long learning cycles.
- Deployment of AI must yield measurable ROI within three months; otherwise, the provider is considered unsuitable for the customer.
- Operational costs are significantly lower than competitors at $3 per million tokens versus $125 per million tokens, with 40 times less CO2 emissions generated.
- Strategic development will rely on open-source foundations like Nemotron and French models, with value driven by reinforcement learning and post-training rather than base architecture.
- Computer use is identified as a key solution for legacy systems lacking APIs, enabling ROI creation in a couple of days instead of months.
- Organizations are expected to face a choice between automating existing workflows or fundamentally changing business models to unlock human intention.
- CEOs of large organizations increasingly view the lack of specific execution capabilities as a critical insufficiency in current technology investments.