Ever wondered what news would look like if it was just for you?

What if the New York Times Application lets users “watch” or “listen” to the news based on how much time they have and how they feel?

What if the New York Times Application lets users “watch” or “listen” to the news based on how much time they have and how they feel?

Role

Role

AI Experience Architect

AI Experience Architect

Design Technologist

Design Technologist

Responsibilities

Responsibilities

Preserved editorial integrity

Preserved editorial integrity

Prototyped AI-assisted editorial experiences

Prototyped AI-assisted editorial experiences

Mapped system behaviour

Mapped system behaviour

Tools

Tools

Figma & Figma Make

Figma & Figma Make

V0 (Vercel)

V0 (Vercel)

Cursor

Cursor

Timeline

Timeline

Jun- Sept 2025

Jun- Sept 2025

OVERVIEW

OVERVIEW

A Next Generation “For You” tab that introduces a personalized news experience which adapts to the format, pace, and context of the reader. It enables discovery and control while reinforcing the Times’ editorial voice and integrity.

Breakdown of the features

Breakdown of the features

  • Dynamic reading modes aligned with the reader context and attention

  • Interest-based filters shaping your feed that balance relevance

  • Adaptive article length based on user-selected time constraints

  • Voice prompt-driven discovery for surfacing relevant articles and quick updates

I designed the Onboarding to give readers agency to shape the “For You” tab’s behavior.

An immediate pop up, allows readers to intentionally choose how they engage, selecting content type, depth, and time commitment before reading.

We mapped news experiences to counteract the silos created by personalization.

BACKGROUND

BACKGROUND

BACKGROUND

BACKGROUND

The need for personalization  beyond echo-chambers

The need for personalization  beyond echo-chambers

The need for personalization  beyond echo-chambers

In a high-choice media culture, readers increasingly rely on personal relevance to navigate information without becoming overwhelmed.

1

Audience

Preference shift in formats based on speed and synthesis based on the user

2

Content

Users want “Related but New” feeds- they seek relevance and surprise while navigating news platforms

3

News form

Delivering digestible content while conserving the tone and context through user-controllable multimedia formats

4

Integrity

Reinforcing editorial voice through specific use cases like Cooking, Games, The Daily, The Athletic, etc.

Experience Pillars of the NYTimes App (Existing)

Experience Pillars of the NYTimes App (Existing)

Experience Pillars of the NYTimes App (Existing)

Author Importance

Long-form articles, explainers, and investigative journalism with clear attribution and emphasis on editorial integrity

Use of AI

Used to extend accessibility and suggestions (location, history, and interests), but limited to help feel the experience is intentionally sectioned

Language and Media

Not a Mega App, it is editorially curated, and not algorithmically driven

Non-News Zones

Wordle, Spelling Bee, and Cooking drive daily return habits and outperform time-spent per session

PROBLEM

PROBLEM

PROBLEM

PROBLEM

Fragmentation and polarization lower satisfaction and weaken trust, undermining the credibility of traditional news media.

  • No preview into the format of the content except the title and description

  • No clear filtering for the content type

  • Lack of options to explore during the journey

  • Lack of visual cues or illustrations to quickly understand a topic

  • No onboarding or visual hierarchy

  • Discovery is difficult due to the layout

  • Landing page logic is unclear. Is it trending, personalized, or random?

  • Feels static and hard to explore

  • Just “one article at a time.”

  • Little user control over visible topics or genre filters

  • Categorization logic isn’t transparent or adjustable

AIM

AIM

AIM

AIM

By offering summarized, digestible, and transparent content ( without changing the content), NYT can reach these groups without compromising editorial standards.

By offering summarized, digestible, and transparent content ( without changing the content), NYT can reach these groups without compromising editorial standards.

By offering summarized, digestible, and transparent content ( without changing the content), NYT can reach these groups without compromising editorial standards.

RESEARCH

RESEARCH

RESEARCH

RESEARCH

"Personalised Diversity" promotes the idea of exploration beyond familiar viewpoints, thus still enabling relevance.

For eg, " Surprise me", " Related but new'

"Personalised Diversity" promotes the idea of exploration beyond familiar viewpoints, thus still enabling relevance.

For eg, " Surprise me", " Related but new'

"Personalised Diversity" promotes the idea of exploration beyond familiar viewpoints, thus still enabling relevance.

For eg, " Surprise me", " Related but new'

Competitive Audit

Competitive Audit

GOALS

GOALS

GOALS

GOALS

High level Vision

Users get articles that are consumable by them to ensure users spend more time with the content.

User & Business

Increase engagement across the entire NYT platform, not just within isolated content zones

Planning

Implement a recommendation engine that follows the editorial standards

Competitive Audit

Competitive Audit

HOW MIGHT WE

HOW MIGHT WE

HOW MIGHT WE

HOW MIGHT WE

How might we help users navigate news that is catered to their availability, surroundings, and emotional state without compromising the essence and depth of journalism?

Experience Framework of the NYTimes App (Proposed)

Experience Framework of the NYTimes App (Proposed)

Experience Framework of the NYTimes App (Proposed)

IDEATION

This wasn’t about more content, it was about fit.

This wasn’t about more content, it was about fit.

This wasn’t about more content, it was about fit.

News that fits a user’s time, intent, and preferences, and can be consumed in different ways without losing context or credibility.

News that fits a user’s time, intent, and preferences, and can be consumed in different ways without losing context or credibility.

SCENARIOS

1

Intent focus: Mood and mode based filters

“I want to sit back and check out some relaxing, fun picks”

  • This choice signals intent and cognitive state before the system ever generates content

  • Each mode represents a system state, not just a filter

2

Onboarding and Prompting

“What is the You Tab?”

3

AI voice prompting and summarization

“I have 10 minutes and I want to know about some wildlife highlights today”

4

Multimedia Exploration

“I am taking a break from work and want to get a glimpse of this saved audio piece within 5 mins as I scroll through other stories.”

This scenario explores how we designed AI-assisted multimedia exploration for moments of low attention, without breaking context or control.

Objective

  • We don’t aggressively personalize

  • Users communicate naturally

  • We do not lose them at the preview of features

  • AI systems need to adapt to user intent

  • No loss of cognitive state

Feature

  • GPT Summary within Journalism Boundaries

  • Text to Speech (TTS)

  • No popups for onboarding

  • Dynamic content feed generation

  • Microinteractions: emotive UI

ARCHITECTURE

ARCHITECTURE

ARCHITECTURE

ARCHITECTURE

How and where AI would play its role?

DELIVERABLE

We landed on 5 main features that decreased clutter but increased agency of the user.

A Next Generation “For You” tab that introduces a personalized news experience which adapts to the format, pace, and context of the reader. It enables discovery and control while reinforcing the Times’ editorial voice and integrity.

A Next Generation “For You” tab that introduces a personalized news experience which adapts to the format, pace, and context of the reader. It enables discovery and control while reinforcing the Times’ editorial voice and integrity.

Overview

Modes

Constrains

  • Retain references and the content

  • Integrate User patterns with respect to history, location and preferences set initially

  • User understands the navigation and has the flexibility to change it

Voice Prompting

Define articles based on your time

LEARNINGS

Designing a governed system under strict trust, accuracy, and accountability constrain using human agency

One solution should not fit all - But it should be aligned to maintain consistency

Future Changes

Check out other projects

Directed by,

Trisha Sharad Pawar

trishapawar23@gmail.com

+1 (206) 931 0943

Check out other projects

Directed by,

Trisha Sharad Pawar

trishapawar23@gmail.com

+1 (206) 931 0943

Check out other projects

Directed by,

Trisha Sharad Pawar

trishapawar23@gmail.com

+1 (206) 931 0943