01 — AI product design
Generation with boundaries
AI writes the content; the template keeps layout, hierarchy and interaction. Every generated surface is a draft the user reviews, edits and publishes — nothing goes live on its own.
Status · available for product design work
Complex problems.
Clear systems.
I’m Pragya, a Product Designer simplifying complex SaaS products across AI, learning, platforms and business workflows.
What I actually do
Five kinds of complexity, and what changes when each one is treated as a system.
02 — Real-time systems
One lifecycle, two roles
Joining, device readiness, floor control, permissions, moderation, recording, disconnection and recovery — designed once as a single lifecycle, for the instructor and the learner at the same time.
03 — Platform · design systems
Frameworks, not screens
Reusable admin patterns across Settings, Create and Report workflows, replacing one-off screen design. Recurring interface work went from about a week to about an hour.
04 — Information architecture
One structure, three expressions
Courses, community and conversation as one navigable system — a configurable desktop sidebar, a native app, and mobile web, all reading from the same architecture.
05 — Commerce · conversion
Where business rules meet the UI
Offers, coupons, checkout, entitlements and promotional surfaces, where the constraint is rarely the screen and almost always the rule underneath it.
Selected work
A few products where complexity came from different places — real-time interaction, AI, platform architecture, business workflows and learner behavior.
01 — AI Product Design · Activation · Personalization
From a blank LMS to a launch-ready school in minutes
Existing setup overwhelmed new teachers: too many manual steps, and no clear first thing to create. Every generated output stays reviewable and editable.
Learnyst — AI onboardingFictional schools · synthetic content
Tell us about your teaching
What do you teach?
Who do you teach?
How do you teach?
Generated · draftNothing published yet
Nayan Design LabDesign · cohort based
Course
Product Critique Intensive
12 weeks · live
Sales page
Learn design by shipping, not by watching.
Enrol now
App
Nayan
Target — not achieved
< 10 min
Average onboarding time
A product target the work was designed against, not a measured production result.
02 — Complex UX · Real-time Systems · Web + Mobile
Designing a classroom, not another video meeting
Device readiness, joining, instructor and learner state, chat, Q&A, raise hand, polls, moderation, permissions, recording, rejoining and failure recovery — one lifecycle, designed for both roles.
Outcome
NPS 7
Live classes
SuperLive — Product Design CritiqueFictional session · synthetic learners
Priya NairInstructor · holding the floor
StageRecording
AaravListening
MeeraListening
RohanListening
SnehaListening
KabirListening
DivyaListening
ImranListening
TaraListening
Raise-hand queue0
No requests. The floor is Priya’s.
Eight learners are in the room. Priya holds the floor; nobody else has a microphone.
03 — Platform Design · Design Systems · Developer Efficiency
~7 days
~1 hour
Recurring admin interface work, for supported framework cases. An actual operational outcome, not a projection.
Turning a week of SaaS UI development into about an hour
Reusable admin patterns across Settings, Create and Report workflows replaced one-off screen design: requirement → shared framework → configuration → reusable interface. The platform already used a metadata-driven, low-code admin architecture; the design work was the framework and its migration patterns.
04 — Information Architecture · Community · Responsive UX
Turning courses, communities and conversations into one navigable system
Educators often relied on WhatsApp, Telegram and Discord outside the LMS, so learning and community became fragmented.
One information architecture
→ three platform expressions
Desktop — a configurable sidebar holds navigation; spaces and their content share one canvas.
05 — AI · Conversion · Template Systems
Giving AI freedom over content — not the design
AI replaces structured textual content only. Markup, layout, component hierarchy, styles, responsive behavior and dynamic widget contracts stay with the template.
Course sales page — generatedFictional course · synthetic prices and reviews
Cohort · starts 4 March
Learn product design by shipping, not by watching.
Twelve weeks where every module ends in a critique with a working designer. Built for people who already have a job and no patience for theory.
Pricing · from the course record
₹4,999 ₹7,999
Duration12 weeks
FormatLive cohort
Seats left9 of 24
Syllabus · from the course record
Framing the problem4 lessons
Interaction and state6 lessons
Systems and reuse5 lessons
Critique and defence3 lessons
Reviews · 4.6 from 318 ratings
★★★★★
The Friday critiques were uncomfortable in the way that actually changes how you work.
Aarav Shah★★★★★
I came in able to make screens and left able to explain why they were those screens.
Meera IyerInstructor · from the author record
Priya Nair
Questions
Do I need a portfolio to start?No. You will have four pieces by the end, and the critique sessions are where they get better.
What if I miss a live session?Every critique is recorded, and you keep access after the cohort ends.
AI rewrote this The template and the course record own this
Actual outcomes
35%
Adoption of AI generation
15%
Edited after generation
Product target — not achieved
≥70% generation → publish
Generation-to-publish conversion
A product target set for the feature. It has not been reached.
Some product visuals have been simplified or reconstructed to protect confidential platform and customer information.
More systems I’ve worked on
Range, not repetition. Each card carries the shape of the system.
Turning course completion into a journey learners can see and followStructured learner progression through milestones, sequential unlocking, rewards and completion rules.OpenOne poll. Many learning moments.A reusable engagement object used across live classes, courses, community and other learning surfaces.OpenMaking the best available offer visible before paymentReducing friction between purchase intent, available offers, payment and access.OpenMaking a website builder start closer to finishedHelping educators publish professional websites without needing design expertise.OpenOne search box across a fragmented learning productMaking multiple content types discoverable as one searchable learning system.OpenGiving admins promotional control without giving them a design toolA configurable promotional surface supporting publishing, placement, responsive settings and performance reporting.Open
Scroll the row — 6 projects
How I approach complicated products
01
Find the constraint
Find the real constraint
Is the complexity coming from users, permissions, legacy systems, business rules or scale? Answer that before touching the UI.
02
Understand the system
Design the system, not the screen
Look for reusable behavior and shared states before treating every requirement as a separate page.
03
Reduce complexity
Give automation boundaries
AI is useful when it removes repetitive work without taking away visibility, control or trust.
04
Ship with engineering
Stay through implementation
Design doesn’t end at Figma. I work with engineering through implementation, migration and design QA until it ships.
A designer who likes messy products
I’m Pragya Kumari, a Product Designer working on B2B2C SaaS. I enjoy problems where the challenge isn’t making one screen prettier, but understanding how many roles, rules and systems fit together — then making that complexity feel simple.I started with visual craft. The problems gradually became larger: first screens, then workflows, then reusable systems, and now AI-assisted products operating across entire platforms.More about me →
Visual
2022
Visual / Communication Design
Screen
2023
Visual Design Intern — Learnyst
Flow
2023–2024
Product UX and SaaS workflows
System
2025
Real-time learning and engagement systems
Platform / AI
2026
AI products, design systems and platform architecture