Edison IA Raises €1M and Promises AI ROI in Six Months

The Nantes-based company Edison IA has announced a Seed round of 1 million euros to accelerate a subscription model that combines AI agents, business integrations, and a contractual promise of return on investment within six months. The offering targets SMEs and mid-sized enterprises (ETIs), featuring a multi-model platform and a method for measuring economic gains in an already crowded competitive landscape.

A Measurable Bet: A Six-Month ROI Guarantee

Edison IA markets its Lightning Program at 50,000 euros excluding tax per year, following a training and diagnostic period, and offers a possibility of termination at no cost if the return on investment is not observed within the first six months, in accordance with the contract terms. The company thus claims to take on part of the risk associated with implementation. However, evaluating ROI requires linking time saved to cost reductions or operational optimizations, such as preparing a quote ten times faster, reducing research time by several hours, or automating an administrative process. For this type of gain to be considered a return, it must manifest as increased output, improved service, reduced errors, avoided hiring, or decreased external expenses. The robustness of the guarantee will rely on indicators determined with each client, including the starting point, volumes processed, time valuation, data access, adoption, and exclusion criteria. Edison IA will need to convince clients of the rigor of its evaluation method in addition to its agents and prove that the observed benefits extend beyond pilot stages, especially since SMEs and ETIs mobilize processes, data, and teams far beyond merely purchasing a license.

Connecting AI to Real Work via the Edison Hub and Agents

The solution is based on the Edison Hub, a platform designed as a connection interface between AI models, knowledge repositories, and internal software, aiming to integrate existing models into users' daily routines rather than creating a proprietary model. Three categories of agents are available: conversational agents to assist employees or respond to clients, autonomous agents that execute task sequences according to defined rules, and application agents that use data in a dedicated interface to generate analyses, indicators, and dashboards. An agent is capable of querying data, applying rules, and initiating actions in business tools. Edison IA favors a multi-model approach, utilizing ChatGPT, Claude, Copilot, or others depending on the usage context, security requirements, and budget. The ability to avoid reliance on a single provider is highlighted as a commercial advantage, given the rapid evolution of models. The value of the offering lies in its implementation: software connections, access rights, document quality, validation rules, and team adoption are considered prerequisites for large-scale deployment.

Increasing Usage but Scaling Up Remains Challenging

The use of AI is progressing in French companies, but remains strongly correlated with the size of the organization: according to Insee, 31% of companies with between 50 and 249 employees and 58% of those with 250 employees or more reported employing at least one AI technology by 2025. Transitioning from the prototype stage to deployment is also complex in large organizations. The AI policy of MAIF emphasizes structuring use cases, managing data, controlling access rights, defining control rules, and the ability to switch providers. Many leaders continue to question how to get started, which data to connect, and how to demonstrate that time saved translates into concrete economic benefits.

Target Audience: Between 50 and 5,000 Employees, About 25,000 Companies

Edison IA targets companies with 50 to 5,000 employees, which represents nearly 25,000 firms in France. These organizations generally have an information system, support functions, and recurring processes conducive to cross-functional projects, but do not necessarily have data or AI teams capable of managing them internally. While obtaining licenses for tools like ChatGPT, Copilot, or Claude is immediate, transforming quote preparation, customer service management, planning, document control, or oversight is a complex project due to data dispersion, diverse access rights, and the need to ensure verifiability. Edison IA positions itself between general-purpose tools and business processes, through an annual program combining cultural education, diagnostics, prioritization, agent deployment, and impact measurement. The company is based in Nantes and was founded by Anaïs Vivion and Julien Hervouët.

A Crowded Market Where Edison IA Seeks an Intermediate Position

The competitive landscape ranges from model and office suite providers like Microsoft, OpenAI, and Anthropic to integrators and IT service companies partnering with major clouds, as well as consulting firms, automation agencies, and AI agent specialists for supporting use cases and pilots. The DGE and Hub France IA list 88 solutions suitable for SMEs and ETIs in a 2026 catalog, and Hub France IA's mapping identifies nearly a thousand French players. While many publishers charge licenses per user, API, or integration projects, Edison IA aims for an intermediate position with a subscription that includes a catalog of agents, a presence with teams, and the facilitation of a network of leaders through the Edison IA Club.

Goals, Resources, and Scalability Conditions

With the proposed pricing, convincing 100 client companies would represent a gross recurring revenue of about 5 million euros, a milestone that Edison IA aims to reach by 2028. The team consists of about ten people and plans to expand its agent offerings while recruiting for deployment and business development roles. Profitability will depend on the ability to share connectors, use cases, and methods specific to each sector; conversely, integrations requiring too much expertise would bring the platform closer to a service activity. Edison IA reports a fundraising of 1 million euros in Seed from Go Capital, with support from CIC Ouest and Bpifrance. This funding aims to strengthen development and sales teams, enrich the Edison Hub catalog, and support a national expansion of the model.