Health360 AI Customer Chatbot

Problem Statement

Health360 Health Supplements was experiencing a sharp rise in customer interactions across multiple channels (Call, WhatsApp, Email) with only 5 support agents. This led to overwhelmed staff, delayed responses, inconsistent service, and potential brand erosion. The team needed an intelligent solution to deflect routine queries while maintaining compliance with health supplement regulations.

My Role

End-to-end product ownership from initial concept to production deployment. Responsible for market research, stakeholder management, requirements gathering, solution architecture, development oversight, and performance optimization.

Timeline

September 2025 - 6 weeks from ideation to production

Stakeholder Interviews

• Conducted interviews with CEO, Customer Support Lead, and 5 support agents
• Identified peak interaction times across UK and India time zones
• Analyzed current ticket volume and response time metrics
• Documented pain points and workflow inefficiencies

Market Analysis

• Researched health supplement industry compliance requirements (FSSAI/UK ASA)
• Evaluated existing AI chatbot solutions and identified gaps
• Analyzed competitor customer support strategies
• Determined that generic AI solutions were unsuitable due to compliance risks

Requirements Gathering

• Defined functional requirements: query classification, dynamic responses, escalation workflows
• Established non-functional requirements: 99.5% uptime, <3s response time, regulatory compliance
• Created user stories for customers, support agents, and administrators
• Documented success metrics: 40-60% deflection rate, <2min response time

Technical Architecture

• Designed RAG (Retrieval Augmented Generation) system for accurate product information
• Implemented intelligent query classification (Complaint vs Product Query)
• Created dynamic response engine with approved knowledge base
• Built escalation workflow with confidence scoring (<80% triggers human handoff)

AI Strategy

• Selected appropriate LLM for health supplement domain
• Designed compliance guardrails to prevent medical claims
• Implemented brand voice customization for consistent messaging
• Created timezone-aware response system for UK/India regions

Compliance Framework

• Established content review process with legal team
• Implemented response filtering for regulatory compliance
• Created audit trail for all customer interactions
• Designed fallback mechanisms for sensitive queries

MVP Development Approach

• Phase 1: Core chatbot with basic query handling (Week 1-2)
• Phase 2: Advanced AI features and analytics dashboard (Week 3-4)
• Phase 3: Compliance features and performance optimization (Week 5-6)
• Iterative testing and refinement throughout development cycle

Cross-functional Collaboration

• Coordinated with development team for technical implementation
• Worked with compliance officer for regulatory approval
• Collaborated with customer support team for knowledge base creation
• Managed stakeholder expectations and provided regular updates

Testing & Quality Assurance

• Conducted extensive testing with real customer queries
• Performed compliance review of all automated responses
• Load tested system for 10,000+ queries/day capacity
• Validated escalation workflows with support team

Business Outcomes

• Reduced customer support costs by 40% through automated query handling
• Improved customer satisfaction with instant 24/7 support availability
• Enhanced agent productivity by focusing on complex, high-value interactions
• Achieved zero regulatory violations through robust compliance framework
• Provided actionable insights through real-time analytics dashboard

Key Insights

• Compliance First: In regulated industries, building compliance into the core architecture is crucial rather than adding it as an afterthought
• Stakeholder Alignment: Regular communication and expectation management across all stakeholders prevented scope creep and ensured smooth delivery
• Iterative Approach: MVP methodology allowed for rapid validation and course correction based on real user feedback
• Data-Driven Decisions: Establishing clear metrics from the start enabled objective evaluation of success and areas for improvement

Future Enhancements

• WhatsApp Business API integration for automated messaging
• Email parser for automatic ticket creation and responses
• Voice bot integration for call center automation
• Multilingual support for global expansion
• Advanced analytics with predictive insights