Hi, I'm Utkarsh, a customer-focused builder crafting products at the intersection of technology, data and design.

Developed a no-code agentic workflow in Google Opal to simulate high-stakes triage interviews using multimodal inputs, orchestration agents, and structured evaluations.

I iteratively refined prompt-driven workflows while working through non-determinism, debugging, and system design tradeoffs. The project also surfaced real risks around bias, over-standardization, and AI dependency in hiring.

Designed an AI insights engine that helped product teams surface customer needs, market signals, and decision-ready recommendations across fragmented banking data.

The concept won first place at the Tepper Tech Innovation Challenge and became a strong example of how I like to work at the intersection of product strategy, enterprise AI, and storytelling.

Built a recommendation experience for CRM agents that surfaced next-best actions using workflow context and LLM-assisted reasoning.

Conceptualized and prototyped a vaccine distribution workflow to support citizen registration, appointment scheduling, site operations, and rollout coordination during a ServiceNow hackathon.

Collaborated with three engineers, scoped the MVP by researching public-sector rollout requirements, and deliberated on tradeoffs across registration, inventory, eligibility, and check-in flows. The concept ranked in the top 10% of the hackathon, validating the business opportunity. The company later productized a vaccine distribution app for large-scale public deployment for NHS Scotland.

Built an Android health-support app selected by the IIT BHU Design & Innovation Hub to help students monitor mental-health signals, complete PHQ-9 style assessments, and access a conversational chatbot experience.

The two-month project combined Android Studio, Firebase, and Dialogflow to support sign-up, data capture, offline sync, comparative analysis, and a chatbot named Neo designed to make the experience more approachable.