Evergreen Broker Plan Guide Conversational AI Assistant

Broker Plan Selection Tool

June–Aug 2016

An AI-powered plan guide selection assistant for broker experience. The assistant is capable of email, plan guide selection assistance and explanation activities through text message conversations and with smart assistants.

As Product Manager, I led the creation of Evergreen Health’s AI-powered conversational assistant from concept to launch. I partnered with engineering, design, sales, and executive leadership to define the product vision, conduct broker interviews, and gather market insights. I managed customer onboarding, oversaw AI and NLP training, and translated broker and business requirements into functional product specifications. In my technical lead capacity, I also owned product roadmapping, wrote conversational copy, and ensured regulatory compliance for all digital outputs.

The Broker Plan Guide Assistant is an AI-powered chatbot designed to streamline the plan guide selection and explanation process for insurance brokers. Built to client requirements and broker workflows, the assistant uses Natural Language Understanding (NLU) to interpret everyday language, a Dialogue Manager to guide the conversation and retrieve the right plan details, and Natural Language Generation (NLG) to deliver clear, human-like responses.

Integrated with Salesforce and the broker management platform, it instantly recognizes the broker, their historical sales patterns, and account information—no downloads, no extra logins. The assistant provides end-to-end digital support, from answering plan questions to sending follow-up documentation, freeing sales teams from repetitive inquiries and enabling brokers to work faster and more efficiently.

KEY RESULTS

• 84% adoption rate, with most plan guide and explanation actions performed via the assistant over traditional methods.
• 92% engagement from top users, with some brokers downloading 5× more plan guides per month than average.
• 10,000+ data points processed from Salesforce, conversations, and user activity.
• Seamless omni-channel experience deployed across six integrated platforms.

I designed the machine learning layer wireframe that defines how the assistant processes requests, retrieves accurate plan information, and personalizes responses. This framework mapped data flows from Salesforce and the broker management platform into the ML/NLP model, established classification rules, and outlined branching logic for both common and complex broker queries. By producing this wireframe early, I created a clear technical blueprint that accelerated development, ensured compliance alignment, and optimized the assistant’s ability to deliver relevant plan guidance in real time.

• Lead cross-functional product development for AI/ML-enabled tools.
• Apply conversational interface design principles to optimize user experience.
• Integrate data-driven insights into ongoing product improvements.