Puzzmatic is a domain-specific chatbot platform that offers digital companionship, provides expert business insights, delivers personalized psychological support, furnishes legal assistance, and more through chat and voice bots.

I took charge as the Product Designer. Starting from scratch, I established the branding and developed the color palette, logo, typography, and image libraries. I conducted research, ran experiments, and developed concepts while designing wireframes and prototyping.

Working from scratch in a startup gives you the unbelievable superpower of being a versatile specialist! It's like being in a fast-paced business school where you're both the teacher and the student. There are opportunities to learn, unlearn, and re-learn at every nook and cranny.

If additional funding had been available, I would have prioritized growing our design team and conducting further user interviews.

I learned to see failure as an opportunity for growth and iteration and consider it a valuable career perspective.

I would repeat it at least three times, but I learned by heart how to prioritize (in every area of my life)!

Now I know how to wear the cap of the business owner and communicate with developers! Thanks,Tom Greever, for helping me hone my skills.

When I joined Puzzmatic, significant inconsistencies were undermining our UI design. 1. Our developer created a style guide, pieced together from random design systems and tailwind elements. There was a lack of cohesiveness for an understandable and user-friendly interface.

2. The mismatched design elements did not resonate with the psychology of the product. Our product's appearance seemed disjointed and occasionally unsophisticated, hampering user engagement.

3. There were many design styles - more than 15 colors and six fonts. This overabundance failed to achieve a uniform visual language.

4. The absence of documentation for using the style guide made it challenging to maintain continuity in design updates and work efficiently.

These issues undermined the user experience! It was clear that we needed to create a consistent design system.

Analyze and address inconsistencies in the user interface.

Re-evaluate and align the design elements with the holistic product's psychology.

Simplify the design styles by cutting down the amount of colors and fonts.

Draft and produce clear documentation for the style guide to simplify usage and maintain continuity.

Streamline the user experience to make it more intuitive and user-friendly.

Boost workflow productivity by normalizing the design process.

Build a consistent brand identity that resonates with the product's mission and vision (logo, font, colors, spacing, grid, layout).

As I mentioned above, it's a startup where my boss is my husband, which created some challenges. Also, it's a project where most things are not theoretical, as taught in the academy. So the first, most important, and most challenging stage was to hold meetings, learn to hear, explain, and brainstorm! One of the most common misconceptions of our team (may my husband forgive me) was the False consensus effect. That is where interviews and usability studies come in handy because every conversation must be rooted in data. Eventually, we organized a lot of internal rules and feedback, cross-referenced it with user feedback, and prioritized design changes.

We conducted extensive user research throughout the product lifecycle, collecting a wealth of data. This has provided valuable insights into our users' experiences, including their concerns, values, and goals for our product. In the initial phase of the product, I conducted a Guerrilla user interview (asking about the current version of the website) that gave us insights. I chose this method because it is a great way to gather research insights from people within a shorter timeline and a tighter budget.

While conducting the interview, I used a note-taking method of information gathering; further I collected, grouped, synthesized, and analyzed the information in miro.com

On the basis of the feedback from our guerrilla interview research we created our design system.

Usability and Navigation: Users complained about difficulty finding specific pages (like the personas page), missing browse button, and unclear click targets (we even got notifications about the rage clicks). Online navigation should be intuitive and seamless.

Design Elements: There's feedback about the text being too small for dyslexic users and the chat window being too small on desktops. Users also found the color palette boring and mismatched with the product's purpose, and a variety in font styles was pointed out as confusing.

Chatbot Interaction: Users asked for a snapshot of the in-chat experience (Home page) before they dive in. They also shared their experience of one-sided chat, taking too long for responses, and leaving them uncertain about waiting times.

Pricing Structure: There appears to be no incentive for users to choose a higher plan, indicating that the benefits for different tiers might need more clarification. Also, I applied the Restorff effect.

AI Personality and Customization: Users expressed a desire for more character customization; they also want intimate colors for the AI girlfriend and more details about the personas (such as description, tone, and style). One participant also suggested the possibility of creating their persona.

Payment and Verification Systems: Issues about needing a confirmation page after successful payment and glitches in the email verification process were raised.

Information Presentation: Participants suggested simplifying the homepage and adding guides or instructions for using the chat.

‍Sharing Feature: People are interested in sharing personas with friends, suggesting users would appreciate a social sharing feature.

As I mentioned previously, we created this product when chatbots were not yet hyped, and our main competitors were Chat-GPT, Character.ai, and some small startups, so when analyzing competitors and their design systems, I also asked for help from already known design systems:

Secondary research involved referring to books and articles from the Internet.

After gathering and analyzing all the data, I scheduled a meeting to present my findings and recommendations for creating our Design System. What was decided to create:

Our developers were well aware that a designer moves faster than an engineer, so especially within the limited budget, we decided to create as simple as possible (keeping in mind research insights) to make it real quickly. We agreed that everything we build should be as inclusive, legible, and readable as possible. If we have to sacrifice elegance - so be it. We were building for needs. This is where I fought my perfectionism to the max while calming myself down:

We agreed to use Figma for design work and Linear as a task manager.

I made a schedule of brainstorms and sprints.

An example of the exercise we used during our brainstorming session was the "How Might We?" exercise.

Here are the five phases of our design sprints: understand the scope of the design challenge, ideate possible solutions, decide on the most viable solution, create a workable prototype, and finally, test that prototype with actual users.

When we started building our design system, we thought of design values that inspire, color palettes that pop, typefaces that speak volumes, and icons that are more than just pretty pictures.

We gathered all principles that feel right to our product and satisfy our users' needs. For brainstorming, we used FigJam. Here are I added our Design Systems principles since they overlap with product design principles:

Universal: Our design needs to work for everyone, every culture, every language, every device, every stage of life.

Human: Comfort and trust are key attributes we wish to instill in our users.

Simplistic: A minimal, well-lit space encourages participation and honest, transparent communication.

Consistent: Following Jakob's law and maintaining consistency across the design can improve user learnability and comfort.

Practical: Our product leans more toward utility rather than entertainment, designed for everyday usage while delivering valuable services efficiently.

Fast: Speed is of the essence in our design. We promise a swift, responsive, and readily available user experience.

Transparent: We value the ideas, thoughts, and conversations our users share, reciprocating their trust with our commitment to absolute honesty and transparency.

Guided by the psychology of color, we chose the Analogy Triad Color palette (three-spaced colors on the color wheel). Analogous colors (those next to each other on the color wheel) create a calming effect, while triadic colors are more vibrant. Like fonts and illustrations, colors impart a tone and emotional balance that affects how our users perceive your brand.

Purple and blue generally evoke feelings of trust, sophistication, and tranquility.

I organized a table with all the colors, including colors for error messages and disabled features, including the names of colors, their HEX codes, RGB, their specific usage in the design, and visual examples from the product to illustrate their application.

Eventually, we got a document with all colors for the guideline, the product, and additional products, as a pitch deck and mass media materials.

Throughout all stages of creating the design system, we went through accessibility checks, including text contrast tests at this stage. As a result of initial checks, we found problematic areas with unreadable colors, but later, we tweaked the colors to make our product inclusive.

We settled on grotesque fonts for headlines and body text, as they are known for their legibility and contemporary appeal.

Both fonts met our design principles and main criteria:

Fonts are easy to read and understand, especially at various sizes and weights. "Outfit" has clear letterforms and spacing for optimal legibility.

The fonts align with the tone and personality of our brand. Since Puzzmatic focuses on digital companionship and expert insights, an approachable, trustworthy, and professional font is essential.

The fonts complement the other design elements, such as our color palette and iconography, to maintain consistency across the platform.

Both fonts meet accessibility guidelines.

An example of our typography documentation:

I've included some of the components in the screenshot from the system design documentation. We used Midjourney prompts adapted for our product to search and generate pictures.

Everything was checked for accessibility.

In my passion for creating a harmonious and cohesive visual experience, I used the 8pt grid system. The core principle to realize is that each element and space of the UI can be a multiple of 8.

Examples of the Sizing and Spacing documentation:

After creating the design system, I designed the mockups, applying all the rules from the documentation. Then, I organized a usability study focusing on checking colors, fonts, and spacing. In the picture below, I have shown the structure of the usability study.

There were 5 participants in the study, and the interview lasted 15-20 minutes. During the usability study, I used a note-taking method, which helped me organize and synthesize all the information on miro.com.

After analyzing and identifying all the insights, I created a presentation for our team with findings and recommendations.

For some users, particularly those with visual impairments, the contrast between the text and the background may not be sufficient, especially with the lighter purple shades. Increasing contrast could hinder readability and accessibility.

Add animations or highlights when a user hovers over an AI persona to signify the ability to interact with them.

Increasing the font size or changing the color of the CTA to stand out more against the background gradient could improve visibility.

Add a modal window for Reviews. This would help users understand there are more reviews to read without having to interact with the widget first.

Including a section that more explicitly outlines the kinds of projects the AI Designer specializes in could help users identify if their needs match the AI designer's skills.

Add similar AI Personas.

Change the header to match the style of other pages.

While the header "The right price for you, whoever you are" is inclusive, it may not clearly communicate the page's purpose. Consider using a more direct header like "Choose Your Plan" to set the immediate context.

The 'Monthly' and 'Yearly' toggle could be styled to indicate more clearly which option is selected. A stronger contrast or a more pronounced switch might improve usability.

The 'Most popular' tag on the 'Adventurer' plan seems floating. Integrating this label (like a corner ribbon) into the card design can make it look more cohesive.

Change the button style according to our design system.

Adding tooltips or a short sentence explaining what 'long-term memory' or 'custom personas' entail could assist users in making more informed decisions.

Organize information more comparably. Clarifying all differences can play a key role in decision-making.

Improving visual cues like adding 'Apply' and 'Clear' buttons would make the filtering process more intuitive.

Add a modal window for Reviews. Allowing users to click and read the full review without navigating away from the page would improve the user experience.

Change the header.

Change the logo from Twitter to X for social login options.

The social login options (Google, Windows, X) could be more pronounced. Visual representation, such as the colors, may encourage users to use these options and speed up the onboarding process.

After presenting my findings and recommendations, we organized a brainstorming session and prioritized changes. I returned to the mockups to apply the test findings to my prototype because even small iterative changes can significantly impact user satisfaction and overall product success (mockups and prototypes passed five usability studies).

As a typical startup at the beginning of his era, we had a few constraints, a common scenario in the real world. But I want to show you this scheme to help you understand that quality work can still be achieved despite limitations.

For a project to be successful, these three factors must be balanced. If you break the Iron Triangle by changing to one constraint, the other two must be adjusted accordingly; otherwise, quality will suffer. But we all did a great job, learned from our mistakes, and will continue doing a fantastic job. What we have now:

Established an organized design system.

Completed essential usability studies, which shed light on several areas for improvement.

Implemented iterative design that's set for future improvement.

Working within restrictions of team, scope, and budget has its challenges, but it offers valuable experience in resource management!