title
Parinaz Kassemi
Product Designer
Click to unlock
Preparing desktop…
Four projects covering a self-built mobile web app, a new product launch, AI product design, and my Figma-to-Framer web design process. Choose one from the side tab to open the full write-up.
A mobile journaling experiment about self-expression, thoughtful consumption, and creating something just because it’s delightful to use.
Daydreamz is a mobile-first journaling app built around one small prompt a day. I designed it in Figma and built it in Lovable as a product experiment — extending the ideas behind my portfolio into a new space, one centered less around my own work and more around giving other people a place to express themselves.
There’s no larger business plan behind it. It’s a thought experiment, built to be used, played with, and enjoyed.
My portfolio was submitted for several design awards, which brought a lot more attention and interaction to it than I originally expected.
It’s designed as an interactive desktop, so there’s already quite a bit to do: people can leave sticky notes, drop stickers, name their cursor, and move through the space alongside other visitors. I wanted visiting it to feel less like scrolling through a collection of work and more like spending time somewhere — with small, delightful experiences built into the space for people to discover.
But ultimately, a portfolio is still a space for me to express myself. Daydreamz started with the idea of extending that experience outward.
A lot of our time online is spent consuming things other people made. I was interested in creating a different kind of interaction: something that could begin with thoughtful consumption but eventually invite you to make, answer, decorate, or leave something of your own behind.
One question a day. Answer it or don’t. Add a photo. Drop a sticker somewhere. Draw something. Save the song you keep playing. Write down what you’re reading or watching. The goal isn’t productivity — it’s expression.
Daydreamz opens with a single question about your life — what song did you have on repeat? what made you laugh today? what are you looking forward to? You can answer it, skip it, or come back later.
Around the prompt is a small space you can make your own with stickers, photos, songs, doodles, a line about yourself, and whatever you’re currently listening to, reading, or watching. Over time, those little pieces start to become a record of what your life felt like — without requiring you to sit down and document all of it.
The interactions are small and casual, so Daydreamz should be something you open whenever you have a thought — not something that requires sitting down at a computer.
Blue gradients, rounded type, a butter-yellow accent, and lots of open space. Connected to the playful world of the portfolio without recreating it. More like looking out a window than opening another dashboard.
No streaks to maintain, no pressure to complete an entry. The language follows: “keep a little piece of today,” rather than “complete today’s entry.” Forget about it for three days and nothing has been lost.
Daydreamz is intentionally a thought experiment. There’s no roadmap toward monetization and no larger plan to turn it into a journaling company. Its purpose is much closer to the portfolio it came from: make something thoughtful, put it into the world, and give people an interesting way to interact with it.
The portfolio begins with my own expression. Daydreamz extends that idea by giving some of that space back to the person visiting it. For someone who finds my portfolio, enjoys spending time there, and wants to keep exploring, Daydreamz becomes another little world they can step into — except this time, it’s theirs.
It’s live, it’s still changing, and that’s really the point.
Try it on your phone →
A look at how I use Figma, generative tools, and Framer to take a website from an initial desktop concept to a responsive, working product.
I’ve designed plenty of websites through client work, but much of that work isn’t something I can share publicly. So through jnpr.labs, I’ve started designing and selling my own website templates instead.
It gives me a place to experiment with how I want to design for the web now: starting with the visual direction myself, using generative tools where they actually make the process faster, and carrying the work all the way through to a responsive site in Framer. This case study documents that workflow.
I usually begin by designing the landing page myself in Figma. This is where I make the decisions I don’t want to hand over to automation: hierarchy, composition, typography, pacing, layout, and the overall visual idea.
I also create the imagery alongside the interface, usually using Midjourney. That lets the imagery and the website develop together instead of treating visual assets as something to drop in once the page is already designed. By the end of this stage, I have the desktop experience and visual direction established.
Once the desktop direction is working, I adapt it for tablet and mobile. Responsive design used to mean manually rebuilding and adjusting every section across breakpoints. I still review those decisions myself, but I’ve started using Figma’s newer AI tools to handle more of the first pass.
The agent can take the desktop design and rebuild it for tablet and mobile, which gives me a responsive starting point without manually reconstructing the same interface three times. From there, I go back through the layouts and make the decisions that still require judgment: what collapses, what stacks, what gets smaller, what needs more room, and what simply doesn’t work on a smaller screen.
Once the layouts are resolved, I move the work from Figma into Framer. Using the generated code as a starting point means I don’t have to rebuild the visual design from scratch — I can bring the work over and spend more of my time making sure the actual website behaves the way the design intended.
From there, I refine the responsive behavior, interactions, spacing, assets, and details directly in Framer until the template works as an actual site rather than a collection of screens.
My current workflow looks something like this:
The interesting part of this workflow isn’t asking AI to design a website for me. It’s figuring out which parts of making a website actually benefit from automation.
I still want to establish the concept, composition, visual language, and hierarchy myself. But once those decisions exist, there are parts of the process that are repetitive: adapting layouts, rebuilding structures, translating designs between tools, and preparing multiple breakpoints. That’s where I’ve found AI most useful.
The result is a workflow where I can spend more time designing the parts that need judgment and less time manually reproducing decisions I’ve already made. And because I’m building these as templates through jnpr.labs, I get to keep experimenting with that process every time I make another one.
Designing an AI concierge for a complex wellness ecosystem.
Life Time had evolved far beyond a traditional fitness app. Members could book classes, reserve childcare, browse articles, shop products, register for events, and access hundreds of services — all within a single experience. As the platform continued to grow, discovering information became increasingly difficult, resulting in high volumes of repetitive questions for the in-house concierge team.
At the same time, Life Time was preparing to launch an AI-powered assistant and expand the app beyond members. I joined the initiative to lead discovery efforts that would help define how an AI concierge could become a more intuitive way for users to navigate the ecosystem.
Although the app offered hundreds of valuable services, they were distributed across many different business units and navigation paths. Members frequently contacted concierge staff with simple questions like:
The information often already existed inside the app — it simply wasn't easy to find. The challenge wasn't creating more features. It was creating a simpler way for people to access the ones that already existed.
Twelve service areas each owned their own corner of the app, and each corner had its own logic for how things were named, grouped, and found. A member with a single question had to guess which business they were asking.
That mapping is what made the case for a conversational layer: not another destination, but a way through all of them.
To understand both user needs and business priorities, I led discovery across multiple groups.
I interviewed concierge teams to understand the questions they answered most frequently, where members experienced friction, and which interactions consumed the most time. This helped identify opportunities where AI could provide immediate value by handling repetitive, informational requests.
Life Time encompasses numerous businesses — including fitness, childcare, retail, events, classes, and editorial content. I facilitated workshops with leaders across these business areas to understand what information their customers searched for, which services they wanted surfaced, and how AI could support their unique experiences.
I evaluated conversational AI experiences, including ChatGPT and digital banking assistants, to understand interaction patterns, conversational structure, and approaches for handling uncertainty and unsupported requests.
The most useful finding was where everyone was weak: almost nothing handled its own limits gracefully. That became the space worth owning.
| Pattern | Coverage across products | Take |
|---|---|---|
| Suggested starter prompts | Adopt | |
| Streaming, conversational tone | Adopt | |
| Deep links into existing features | Differentiator | |
| Explicit "I can't help with that" handling | Gap to own | |
| Handoff to a human | Required |
The research revealed three patterns that shaped every decision that followed.
Using the research findings, I developed the artifacts that shaped the conversational experience: journey maps, conversation flows, intent mapping, prompt libraries, and edge-case scenarios.
L·AI·C didn't exist yet — there was no interface to redesign. The first artifacts were conversations on paper: what a member would say, what a good answer looked like, and where each answer should lead next.
Special attention went to how the assistant should respond when a question fell outside its capabilities — keeping the conversation useful while maintaining trust.
Every dead end resolves into a next step, so a limitation never reads as a failure.
The project established a research-backed foundation for Life Time's AI concierge. By aligning customer needs, concierge workflows, and business priorities, the work helped define how conversational AI could improve information discovery while supporting the long-term goal of reducing repetitive concierge requests.
The experience strengthened my understanding of designing for AI — not as a replacement for traditional interfaces, but as a new interaction layer that simplifies access to complex systems.
This project reinforced that successful AI experiences begin long before interface design. Rather than starting with technology, we began by understanding people: what they asked, where they struggled, and how they naturally sought information. Those insights became the foundation for designing a conversational experience that served both users and the business.
Scaling a growing fitness content library.
Life Time's On Demand platform was rapidly expanding. With a new production studio, the business was preparing to release hundreds of high-quality fitness classes across yoga, strength, cycling, Pilates, meditation, and more. The existing experience wasn't built to support that scale.
Members were presented with a continuous feed of videos, making it increasingly difficult to discover relevant content as the library grew. I led the redesign of the On Demand discovery experience, balancing immediate business needs with a long-term product vision that could scale alongside the platform.
The On Demand experience had outgrown its original design. As hundreds of new classes were added, users were expected to browse a single, continuously scrolling feed to find workouts. While this worked for a smaller library, it became increasingly difficult to discover content as the platform expanded.
At the same time, the organization faced competing priorities. The brand team was preparing a company-wide rebrand, while engineering resources were focused on rebuilding the broader Life Time app. The challenge became finding a solution that improved discovery immediately while creating a roadmap for a more robust experience in the future.
One feed, no shape. Every class occupied the same visual weight, so a member with fifteen minutes and a specific instructor in mind had the same path as someone with no plan at all — scroll until something looks right.
The library was about to more than double. Scroll was the only tool, and it was already the bottleneck.
I began by understanding how members searched for fitness content and how other digital fitness platforms approached discovery.
| Pattern | Coverage across platforms | Take |
|---|---|---|
| Search | Expected | |
| Filter by duration | Adopt first | |
| Bookmarking / save | Adopt | |
| Personalized recommendations | Long-term | |
| Category organization | Phase one |
I conducted interviews to better understand how members chose workouts, what information influenced their decisions, and where they experienced friction while browsing the existing library.
I evaluated leading fitness and streaming platforms to identify common discovery patterns, including search, filtering, bookmarking, personalized recommendations, and content organization. This research helped establish which interactions users already expected and where the current experience fell short.
Three themes emerged consistently throughout the research.
Based on these insights, I proposed a phased product strategy. To quickly improve discoverability, I introduced lightweight category filters that organized the growing content library without requiring significant engineering investment.
This provided members with an immediate improvement while aligning with existing technical constraints — a design that could ship inside the capacity the team actually had.
Together, these features created a foundation that could continue supporting the platform as the content library expanded.
I collaborated closely with engineering and business stakeholders so each recommendation balanced user needs with technical feasibility and organizational priorities.
The project established both an immediate solution and a long-term product vision for On Demand. Rather than waiting for engineering capacity to build every feature at once, we introduced an incremental approach that improved content discovery while creating a clear roadmap for future enhancements.
The experience reinforced the importance of designing products that scale — not only for users, but for the organizations building them.
One of the most valuable lessons from this project was recognizing that good product design isn't always about delivering the ideal solution immediately. Sometimes the best outcome is creating a strategy that balances user needs, technical constraints, and business priorities while moving the product forward incrementally.
I’m a digital designer focused on craft, emerging technology, and making thoughtful digital experiences feel beautifully simple.
I’m naturally curious and a little relentless when it comes to figuring things out. I like finding opportunities, making something out of nothing, and looking for another way when the obvious path doesn’t work. I’m always learning something new — whether that’s keeping up with culture and technology, reading about history and philosophy, or picking up a new creative hobby.
I run jnpr.labs, an independent creative lab for digital tools and experiments, and I’m currently building Daydreamz, a mobile-first journaling app.
When I’m not designing, I’m probably at a studio class, trying a new workout, hiking, swimming, doing yoga, learning music production, or getting an overpriced matcha latte somewhere.
Curious about what guides my work? Open the notebook on the desktop.
I help startups and modern brands turn complex ideas into clear digital experiences — product design, websites, pitch decks, data viz, social, and 3D.
Have a project in mind? Let's talk.
inquire →Strategy, UX, and interface design for startups, SaaS, and AI products.
Marketing and product sites for founders and growing tech companies.
Investor, sales, and product decks that raise money and move ideas.
Dashboards, charts, and infographics that make complex data usable.
Cohesive templates and launch content for founder and product brands.
Custom 3D artwork and visual systems for a distinctive brand look.
Welcome to the interactive desktop. Let me show you around.
Parinaz Kassemi