Everything you actually want to know answered

Bubble Invest builds your AI employee, trained on your processes, and trains your team to work alongside it. Our mission: give you the tools that large models don't build for you, tailored to your business, ready to use from day one.

The Quick Start starts at €2,000 excl. VAT for a custom AI agent, including 3 working sessions. The first quote is built with you based on your actual needs. For more complex architectures (multi-agent, advanced integrations), we provide a tailored estimate.

You pay once, with no subscription or recurring fees imposed. Your data, your configuration, and the content produced belong to you, and your agent runs on your infrastructure. You keep free, ongoing use of it. An optional maintenance package is available for long-term support.

We work with small and mid-sized firms, independents, and regulated professions, with a core focus on wealth management and asset management. That's where AI has the most impact: fewer hierarchical layers, visible results immediately. For other sectors, feel free to reach out.

Absolutely. Beyond the Quick Start, we design bespoke solutions: multi-agent teams, selfware (personal AI tools), dedicated applications, and complex architectures. Each project gets a quote scoped to your needs.

Support is included in the days following delivery to fix any initial bugs. Beyond that, we remain available for quick interventions on a quoted basis, and a maintenance package is offered for ongoing support, tailored to your needs.

Yes. Once the agent is deployed, we offer adapted maintenance contracts: a monthly retainer quoted per project for regular follow-up, or on-demand intervention at €500 excl. VAT/day, with a guaranteed 24h response time, remote or on-site.

On three combined levels. Architecture first: the agent runs on your own infrastructure (your machine or a private secured server), so your data stays with you. Setup next: depending on your sensitivity level, we configure the solution to prevent your data from being used to train any AI model (see the dedicated question below). Framework last: contractual confidentiality and a data processing agreement, with your data returned or deleted at the end of the engagement. What's confidential never leaves your environment.

Primarily a leading frontier model, while remaining flexible based on your preferences. One important point: agents run on your own subscription. You keep full control over your costs and your data.

It depends on the setup, and that is exactly where we help you choose. Rather than a flat yes or no, we guide you toward the right option for your sensitivity level: an enterprise plan configured with no data retention and no training on your inputs; local pseudonymisation that masks sensitive data before any processing; or an open-source model hosted entirely on your own infrastructure. Our role is to find the right balance between performance and protection for your situation, and to help you distinguish governed enterprise AI from consumer AI, which uses your data for training by default.

Your choice: locally on your machine or internal server, or on a dedicated private server in Europe. You retain full control of the infrastructure. No dependency on a third-party SaaS platform is imposed.

Yes, this is a standalone service, beyond protecting our own engagements. It covers pseudonymisation of your data, data-flow mapping and the processing register, Data Protection Impact Assessment (DPIA), an AI usage charter, breach response procedure, and a compliance audit. Built by people who understand both the legal and the technical side.

No endless slide decks, no six-figure quotes, no army of junior consultants. We go further than advice: we actually build and deploy a working AI agent from your data and your tools. A turnkey solution that absorbs your workload from day one, without creating dependency.

The first session is free and commitment-free. After signing, the project starts within 1 to 2 weeks, with full delivery in 3 weeks: a turnkey agent, ready to use on your server or locally.

Simple and transparent: quote and contract signed electronically (Yousign). The project starts upon receipt of the deposit. Payment schedules are available and agreed in advance.

Yes, and it's hands-on: no slides or theoretical lectures, just practical sessions where you learn by doing, directly on your agent. We don't leave until your team is comfortable with it. Think of it as onboarding a new colleague who already knows your processes.

We bring a rare combination: strategic consulting (finance, business processes) and AI engineering. We understand your workflows before we write a line of code. And above all, we use our own agents every day. What we build for you, we rely on ourselves. Faster, more reliable, and more affordable than a freelance developer.

Both, depending on your preferences and the nature of the project. Based in the Paris region, we are available anywhere in France and fully capable of working entirely remotely.

Yes, completely. The first session is free and commitment-free. You can decide not to continue after this first meeting: no cost, no obligation. 30 minutes by video call to understand your processes and see what we can build together.

Book a slot via Calendly. We take 30 minutes to understand your organisation, your current tools, and your main friction points. We then suggest a first agent concept tailored to your situation. The whole thing is run by the founders directly, no sales intermediary.