Generative AI: How Seyna Was Turbocharged by Mozza
Seyna's initial request: accelerate AI adoption and explore new opportunities
Founded in 2018, Seyna builds a modern platform dedicated to insurance brokers. It has established itself among the leaders in the insurtech space and attracted 47M€ of funding from top-tier investors (Global Founders Capital, Allianz, Elaia...).
To navigate through the latest advances in generative AI, Seyna's executive team recognized the need for prospective guidance on the changes already disrupting their sector. Already equipped with several insights, they aimed to explore a maximum number of opportunities benefiting their clients with fresh perspectives.
Goals:
- Not just adapt to ongoing changes but anticipate them to gain a competitive edge and position as a leader.
- Implement this generative AI strategy, which requires extensive exploration, without impacting Seyna's current roadmap and strategy.
How to detect and prioritize the best use cases
The goal was to offer Seyna long-term recommendations on AI opportunities for its clients. In short: how can AI enable faster and more accurate services for policyholders and insurance professionals (brokers, managers, and insurers)?
To address this, the Mozza team immersed themselves in a complex sector where automation remains a significant challenge. Despite high digitalization in insurance professions, human expertise is crucial for addressing recurring issues (policy definitions, claims, matching offers with prospects...).
Gathering Insights for Tailored Proposals
To identify the specifics of the insurance market, we explored its entire value chain by conducting over 20 user interviews. These interviews, conducted with insurers, brokers, reinsurers, and Seyna collaborators, enabled us to gather diverse viewpoints and their daily challenges.
We were then able to identify the challenges that Generative AI can address. A mapping of different use cases specific to the insurtech sector was presented to the team, according to their criticality and implementation cost.
Prioritizing Generative AI Opportunities for Seyna
Among the 7 opportunities identified, we have selected 2 avenues due to their potential ROI and ease of implementation:
- A decision-making co-pilot for the sales teams of insurance broker distributors: to more quickly match prospects with the best insurance solution.
- A tool to assist with the management of unpaid rent insurance claims: to automate the handling of claims.
Far from replacing their expertise, these opportunities primarily aim to support employees with complex tasks.
Leveraging our experience as an independent product builder, we also presented Seyna with concrete takeaways and our latest observations on the AI market in France.
Providing technical recommendations and developing several POCs
We shared with Seyna technical recommendations on the implementation of these opportunities, particularly regarding the choice of the LLM (Large Language Model) based on costs and their technical environment. We outlined the advantages and disadvantages of each scenario, our best practices, and the pitfalls to avoid.
"Mozza's tailored approach allowed us to adjust the mission. We wanted something tangible. Therefore, the mission was to identify very concrete use cases for our clients on generative AI. Then, to dive deep into the matter to validate (or eliminate) certain intuitions and test the opportunities identified through several technical POCs." Stephen Leguillon (CEO)
To materialize our recommendations, we created a test platform consisting of 4 agents. These agents, developed in close collaboration with Seyna's team, allowed us to deliver basic but functional POCs for specific use cases:
- Two support agents in the form of intelligent bots specialized in unpaid rents and health insurance with a RAG system, structured knowledge bases, etc.
- A sales support agent for health products with comparisons of health insurance guarantee tables.
- A data extraction agent from client documents to facilitate claims management.
These POCs enabled us to quickly test solutions based on the opportunities identified, closely aligned with Seyna’s needs. Currently, the Seyna team is continuing to explore AI based on the technical recommendations from Mozza.
Stephen Leguillon
CEO
Meet the team
Cofounder of Cousto, ex COO of Ferpection
Made magic happen for:
Finnt, Sorare, Orus
ex Co-founder & CPO Arthur’in
Made magic happen for:
Seyna, La Collection
GenAI expert, ex-Google
Made magic happen for:
Seyna, Brut
Cofounder of Cousto, ex COO of Ferpection
Made magic happen for:
Finnt, Sorare, Orus
More Case Studies
We collaborate with ambitious clients to deliver innovative products.
