Jointly AI, a UK insurtech company focused on automated brokerage, announced the launch of Jointly AI Broker, an end-to-end autonomous artificial intelligence platform developed for UK personal insurance brokers.
The company describes the product as a fully operational brokerage system rather than a conversational assistant. Rather than simply providing a chat response or directing customers to an insurance company’s website, the platform performs the brokerage process on the customer’s behalf.
It uses advanced voice AI models to contact insurance providers directly, navigate automated phone systems, wait if necessary, collect and compare quotes, and return tailored recommendations to customers without the need for human intervention from the broker.
Jointly AI says brokers using the platform can delegate a process that would normally take hours or days and complete it in about 35 to 45 minutes.
The system operates through a coordinated workflow comprised of five purpose-built AI agents, each assigned to a different stage of the brokerage journey. These agents are connected through an enterprise-grade orchestration layer designed to support reliable operations at scale in regulated financial environments.
The process starts with a brief phone call between the customer and the platform’s receiving agent. During this conversation, the system collects the required information without the need for forms or transfers between departments.
The receiving agent is continuously available and uses a speech AI model designed to provide a natural conversational experience. Through this interaction, research agents independently scan the market, check the qualifications of insurers against the Financial Conduct Authority’s register, and shortlist suitable providers.
The quoting agent then calls the insurance company directly. It can navigate the IVR system, wait for a human representative when needed, reference alternative market quotes, and capture pricing information. Multiple calls can be made simultaneously. Generated quotes are passed to analytics agents powered by Jointly Insurance Instruct v1, the company’s proprietary large-scale insurance-focused language model, which normalizes the information and evaluates each option based on customer-specified priorities.
Once the analysis is complete, the delivery agent will communicate the results to the customer via phone or email. Responses include key recommendations, alternatives and budget-oriented choices, each explained in clear, straightforward language.
United AI demonstrates that the orchestration layer is built to meet the expectations of regulated financial services. Each extracted data point is assigned a confidence score, and when certainty falls below a defined threshold, the system seeks clarification rather than making assumptions.
If the call is disconnected, it will be redialed. If a provider is temporarily unavailable, it will try again later, including outside normal business hours if necessary. All agent activities, state transitions, and system operations are recorded and visible in real time, providing supervisory and auditing capabilities for agents.
The platform is available on a subscription basis and is currently available for early access to select partner insurance brokers. United AI is positioning the product as a tool that can reduce the administrative workload within brokerages, a large portion of which is spent on manual tasks, while enabling customers to receive personalized insurance recommendations faster than through traditional processes.
Alberto Romero, CEO of Jointly, added: “We know AI can handle complex tasks, but is it designed to be specialized enough to handle your tasks? Insurance is complex, regulated and far-reaching. It requires AI designed for it from the ground up.”
