Solea came to us in September at seed stage with a clear idea: scheduling is the core business problem, and AI is the right tool to solve it. They needed a design partner and within 24 hours of our first call, we completed a design test to prove we could move at their pace.

After launch, pest control businesses using the platform started running more appointments per day with the same headcount and reported a 10x ROI on the switch.

Solea signed their first national customer within 6 months, raised a $10M Series A, and we are continuing with them into the next phase.

The AI handles all routing automatically.

Thus, manual appointment creation is an edge case, something that happens when a specific customer request or exception requires human intervention.

When that happens, the system surfaces available technicians ranked by fit and shows how much distance and time the appointment adds to their route for that day. The manager picks from a ranked list rather than guessing.

The typography and color choices were made with one thing in mind: the platform needs to feel embedded in whatever system the client is already using rather than standing out as something foreign. We kept everything deliberately neutral.

Light and dark mode both follow the user's system preference automatically.

The route view is the most information-dense screen in the product.

Every technician, route and appointment are all on one screen at the same time. For operators managing large teams, every pixel of space matters and we worked around it.

We built a retractable navigation specifically for this screen. When a dispatcher needs full focus on the route view, they collapse the nav and get the space back. When they need to move between sections, they expand it.

It is a small interaction that reflects the same design system thinking applied throughout the product. Every component should serve the operator's workflow.

Most field service platforms treat communication as a separate problem. Technicians text managers from the field, decisions get made over calls, and nothing is traceable back to the job it came from.

We built communication directly into the platform and thought carefully about where AI could be woven into that experience rather than bolted on top of it. When a dispatcher sends a message, the AI attempts to answer before the technician even needs to respond. If it can find the information, it does. If it cannot, the technician picks it up and the conversation stays in the same thread as the appointment it belongs to.

The dispatcher can also call the technician directly from the same view without switching tools. Every decision, whether resolved by AI or by the technician, is logged in context next to the job it relates to.

Building a platform that covers scheduling, customer records, communication, and appointment booking means designing across very different use cases.

The risk is that each use case starts pulling in its own direction and the product ends up feeling fragmented.

We established a consistent way to use components early and applied it at two levels.

1/ At the page level, consistent text sizes and layout logic.

2/ At the component level, the interaction patterns show whether you are filtering technicians, service types, or job durations.

The result is a system where you don't need many text sizes, components, or colors because everything follows a coherent logic.

The filter panel is one example of that direction.

When a rep is on the phone with a customer they need to orient quickly.

Is this a new customer or an existing one? What was the last thing that happened on this account? What are they likely calling about?

Switching between tools to find that information while someone is on the line creates delays and mistakes.

We designed this view around one principle: show the most recent and relevant information first without overwhelming the rep with the full history.

Documents, appointments, and activity are all accessible but the screen leads with what is important at the moment.

The rep can read the relationship in seconds and navigate to whatever the customer is asking about without losing the thread of the conversation.

The personal information on the right gives the rep everything they need to have an informed conversation from the moment they open the screen.

When a customer calls to book an appointment, the rep needs an answer immediately. Checking calendars, cross referencing routes, and manually figuring out which technician is available while someone is on the line creates delays and errors.

We designed a booking flow where the rep picks a date and the system filters to that day immediately, showing the AI recommendation for the best technician available. The rep does not have to leave the platform or put the customer on hold. The answer is already there. They confirm and move on.

The AI factors in the technician's schedule, working hours, and how much travel time the appointment adds to their route before surfacing a recommendation.

The industry standard for field service scheduling shows three to four technicians at a time, surrounded by completed appointments that are no longer relevant. Dispatchers spend their day mentally filtering noise just to see what is actually happening.

We designed a view where the full team's day is visible at once. All technicians, all AI-powered routes, who is running late, who has gaps, accessible from one place without opening a single individual profile. The view scrolls dynamically with time so completed appointments drop away and the focus stays on what still matters.

For a team of 150 technicians, the filter lets the dispatcher narrow to a specific group, region, or area in seconds. Tomorrow's schedule is already built before anyone starts work because the AI pre-builds routes based on recurring appointments.

Solea AI is a B2B2C SaaS UI/UX design project for an AI services CRM built for pest control businesses.

Most pest control companies use FieldRoutes for billing, subscriptions, appointments, and invoices. They all live in different places which makes scheduling fully manual. The dispatcher drags each appointment into a time slot one by one and figures out the best route, technician, and the correct job order entirely on their own.

The problems we kept hearing about were:

1/ Technician conflicts

2/ Missed appointments

3/ Routes that made no geographic sense

There were too many parameters to hold in one person’s head at once and get right manually.

Solea came to us at seed stage in September with a clear thesis that: scheduling is the core business problem and AI is the right tool to solve it. Within 24 hours of our first call we completed a design test to prove we could move at their pace.

Sunflower is a B2C wellness app for iOS and Android that helped over 10k paying customers as of launch (we went 0 to 10k paying customers with them within 8 months) overcome addiction using cognitive behavioral therapy (CBT) and guided visualization.

Pixel One served as the B2C wellness and health care iOS/Android design agency for Sunflower's iOS and Android wellness app, turning the onboarding experience and core product features from pre-alpha to 100K monthly active users and 10k paying users (free → paid), secured venture funding, and got into Y Combinator.

Sunflower tracks sobriety through a CBT-based visualisation system where every milestone is a meaningful marker in the user's recovery journey.

We designed celebration screens that feel distinct and personal rather than generic, because most recovery apps treat milestones as checkboxes and users feel nothing.

The design had to avoid the preachy, guilt-driven tone that makes people delete wellness apps. The goal is to give users a moment that actually feels earned, something worth coming back for.

Recovery apps lose users in the first few screens. Download is easy and drop-off is easier. Users arrive having already tried other apps that failed them, wondering if this one is any different.

We added real user testimonials directly into the onboarding flow. Proof that recovery is possible when you are in those first few screens, wondering if this app is just another thing that is going to fail you. The quotes are specific, personal, and placed with intention to build trust.

Users spend up to 4 hours straight talking to Sam about their addiction, and they keep coming back because Sam listens without judgment.

Sam is an AI chat companion inside Sunflower. Instead of the usual guilt-driven recovery app tone, Sunflower built a space that felt human and safe.

Many users credited Sam with helping them quit and stay sober. We designed Sam to feel human. The interface had to feel like talking to someone who gets it, instead of a chatbot running through prompts.

Sam became the most beloved feature in the entire app. The healthcare & wellness SaaS app grew from a founder's side hustle to 100K MAU, 10K paying users, $1M ARR, and YC acceptance.

The sobriety timer shows users how long they have been sober, down to the day, hour, and minute. It sits inside the garden they personalised during onboarding, surrounded by the sunflowers they have been growing.

The longer the streak, the more the garden blooms. Users are watching something they built grow.

Sunflower is built around the idea that every sober day grows your garden, and a relapse costs you something. We designed the field switch animation to make that psychological loop visible and felt, not just understood.

The movement had to feel natural and meaningful, soft timing, nothing gamified or cheap, because a recovery app that feels like a slot machine loses the user's trust immediately.

The goal is to make progress feel real enough that users think twice before breaking their streak.

Collective ($82M raised) needed to launch their LLC product without hiring more ops staff. Pixel One designed the entire self-serve onboarding in 6 weeks and sped up activation time by 50%.

Every new customer required hours of manual document verification. We designed OCR-powered upload flows, AI help panels for compliance questions, and mobile signing workflows. The ops team now handles only edge cases.

OCR extracts business names, formation dates, and addresses from uploaded documents automatically. But users still need to trust what the system found before anything gets submitted.

We designed the confirmation flow around editable fields that surface extracted data clearly. Users review what the system found and correct where needed, rather than typing everything from scratch. The design choice was to make the AI output feel transparent and controllable.

The design choice was to make the AI output feel transparent and controllable. The result is a flow that reduces data entry to the minimum while keeping users fully in control of what gets submitted.

The original signing process required users to download documents, sign them separately, then re-upload. Every extra step was a chance to drop off, and many did.

We redesigned the signing flow as a sequential step-by-step process that works on any device. Phone camera upload or desktop, each step shows exactly what comes next so users always know where they are in the process.

The design choice was to break a complex multi-document signing requirement into single actions, so users never faced the full complexity of what they were completing all at once. The signing workflow eliminated the download-sign-upload loop entirely, which contributed directly to the 50% reduction in time to activation.

Compliance onboarding fails when users upload the wrong document and have no idea why it was rejected.

We designed the upload interface around visual examples that show users precisely what a valid document looks like before they submit anything.

The design choice was to surface verification guidance at the moment of upload rather than after rejection, because catching the error before it happens removes the back-and-forth entirely. The goal is to make users feel confident they are submitting the right thing the first time.

Founders uploading business formation documents have no idea what "articles of organization" or "EIN verification letters" are, they created these documents years ago and forgot about them. Static help documentation wasn't working and support tickets piled up with the same basic questions.

We designed an AI-powered help panel as a solution. Users click "What's this?" next to any confusing requirement and get explanations in real-time. The system answers follow-up questions, explains why specific documents are needed, and tells users where they'd typically find them.

The design challenge was surfacing this help without cluttering the interface, so we positioned it as a persistent sidebar that appears on-demand.