AI Hallucination Detection Tool
An interactive, scenario-based tool that guides data evaluators through identifying when AI agents fabricate information. Progressive disclosure with real-time feedback.
Try the prototype →Hi, I'm
Designer
Crafting intuitive design solutions for complex, data-rich environments — from AI evaluation tools to multi-audience content systems. I bring a human-centered approach and a sharp eye for systems thinking to every project I touch.
I grew up in New Delhi, fell in love with English literature and psychology, and eventually found my way to Boston for my Master's at Northeastern, where I discovered what might be my favorite thing in the world: the moment a complicated idea finally clicks for someone.
That thread has run through everything I've done. Designing evaluation tools for AI annotators at Amazon. Building information architectures that turn complex data into clear decisions. Creating content systems at MIT that served three distinct audiences simultaneously. Each time, the work was really the same — take something layered and intricate, and design a way for people to actually navigate it, not just endure it.
Product Design felt like finding the name for something I'd already been doing. It lives right at the intersection of the things I care most about: user research, information architecture, interaction design, and the quiet satisfaction of a well-designed experience. I'm currently looking for roles where I can bring that same care to the people who use the tools I build.
How I approach every product design challenge.
User interviews, workflow analysis, stakeholder alignment, and problem framing
User flows, information architecture, requirements mapping, and success metrics
Wireframes, prototypes, and high-fidelity designs with iterative review cycles
Usability testing, user feedback, pilot launches, and cross-functional validation
Measure impact with data, gather insights, and continuously refine the experience
A closer look at how I approach complex design challenges.
An interactive, scenario-based tool that guides data evaluators through identifying when AI agents fabricate information. Progressive disclosure with real-time feedback.
Try the prototype →A multi-phase assessment interface with simulated audit workflows, form-based interactions, scoring logic, and progressive task unlocking for AI annotation auditors.
Try the tool →Designed the end-to-end evaluation workflow, information architecture, and decision frameworks for 50+ evaluators assessing AI agent performance across 1M+ data points.
Read case study →Designed a multimedia experience that made complex, culturally-specific content accessible to 45 users with no prior context — through progressive disclosure and multi-modal interaction.
Read case study →Built a scalable content system from scratch serving three distinct user segments simultaneously — reaching 168,000+ interactions through research-driven, data-informed design.
Read case study →Amazon
Designed end-to-end evaluation workflows, information architecture, and quality frameworks for 50+ evaluators assessing AI agent performance across 1M+ data points monthly.
Inkitt
Applied UX principles and data-driven design to optimize landing pages and ad experiences, boosting engagement and conversion rates.
MIT Solve
Built a multi-audience content system from scratch, designing information architecture and user-centered narratives that drove 168,000+ interactions across 5 platforms.
Northeastern University
Designed multimedia experiences for a Global Literature course, applying user research and progressive disclosure to make complex content accessible to 50+ users per semester.
Srishti Publishers
Designed information architecture and content structures for complex material, organizing for clarity, scannability, and accessibility.
Humans of Safe Places
Designed user-centered content with a focus on UX, creating accessible educational resources on safety and well-being.