Tarun Hari

Partnered with leadership to define and prototype MeLo, a Gen AI mobile assistant that coordinates and delegates household tasks for working parents in a 50-50 household split. Designed the conversational/voice interface, mapped user intents, built dialogue flows, and established a design system to balance empathy, reliability, and transparency.

Partnered with leadership to define and prototype MeLo, a GenAI mobile assistant helping 50-50 household coordination. Led the design for MeLo’s conversational and voice interfaces, mapping user intents, dialogue flows, failure states, and safety rails to balance empathy, reliability, and transparency. Built functional low-code prototypes in Figma Make to simulate multi-agent orchestration and accelerate engineering alignment, reducing design-to-dev handoffs by roughly 40%. Authored agent-behavior and prompt-pattern specifications outlining tone, consent, and privacy principles.

AI-driven adaptive learning platform; launched at Bolt Global AI Hackathon. Defined end-to-end AI UX architecture prompt pattern, reasoning chains, and use‑case modeling, building a foundation for personalization and analytics. Shipped working product using Bolt (no-code frontend), Supabase, and Groq LLM; expanded to multi-modal voice-first interface at Eleven Labs; led cross-functional product delivery as PM, designer, and delivery lead.