Redesigning the Website to Reflect Its Innovative Work

Redesigned NetworkOcean’s website using the AIDA framework to clearly convey its underwater data center technology.

Design a website for a company that builds underwater data centers and gives an option to it’s users to reserve compute powers/ GPUs.

List of a few companies operating in the same domain

 to guide users smoothly from awareness to action:

Using this approach, I defined the section order to naturally lead visitors toward engagement and conversion.

 — Emphasizes the key challenges and inefficiencies of traditional data centers.

 — Showcases the outcomes and advantages achieved through NetworkOcean’s innovative approach.

 — Explains the process and technology that make the solution possible.

 — Highlights secondary advantages that further strengthen the value proposition.

 — Presents key details about the technology and current GPU availability, along with a clear call-to-action to book the service.

Since NetworkOcean builds submersible data centers, I wanted the website to convey an underwater environment — dark, calm, and built around deep blue tones that emphasize water’s importance in the system. The dotted wave pattern from the original site was expanded into a micro-dot grid to subtly represent data flow, paired with metallic surfaces inspired by the outer hull of the submersible units.

Initially, I was leaning toward a GPU-inspired aesthetic with circuit board patterns and a neon glow feel.

For my iteration 2, I wanted to test different dot sizes and spacing for visual clarity, I looked for a Figma plugin that could convert PNG icons into dotted patterns. Since none matched my exact needs, I decided to build my own tool using Claude and Lovable. The concept was simple enough to prototype quickly with these AI-assisted coding platforms.

https://dot-pattern-generator.lovable.app/

After the first round of feedback, I decided to move away from the maximalist approach and aim for a cleaner, more minimal design. I explored the idea of using dotted patterns to subtly represent the company’s core technology. These dots, combined with metallic textures inspired by a submarine hull, create a balanced contrast between the digital and industrial aspects of the brand.

NetworkOcean

Context about the problem and reasons for choosing the problem statement

Busy doctors who treat multiple patients daily often struggle to keep track of upcoming appointments and pre-appointment action items such as lab results. Certain procedures depend on these results before the next step can begin. While some doctors rely on assistants, managing schedules, sending reminders, and following up with patients is still largely manual and easy to miss.

This was part of an assignment at the 10kdesigners cohort. I chose it because I’ve seen this challenge firsthand : my sister, a dentist, often spends her Sundays planning the week ahead, calling patients to confirm appointments and request reports. I wanted to explore how thoughtful design and AI could reduce this load and improve care coordination.

Through conversations with my sister (a practicing dentist) and her colleagues, I mapped out the typical flow of how treatment planning is handled in real-world clinics:

Tasks like sending appointment reminders, follow-up messages, and post-treatment caution notes take up valuable time. Doctors often manage these manually or through assistants, making it harder to stay organized, especially when handling many patients and frequent changes to schedules or treatment plans.

Communicating updates individually to each patient is time-consuming and prone to delays or missed information.

Doctors need to mentally keep track or manually check where each patient is in their treatment plan, which adds unnecessary load.

Pre-visit requirements like lab results can be forgotten if not tracked properly, leading to treatment delays.

Rescheduling a step or changing a treatment plan requires manual coordination with the patient, which adds effort and can cause confusion.

Before starting the design, I outlined the full range of features the app could support and categorized them into must-haves and should-haves. This helped me focus on the core workflows doctors need most - like scheduling, reminders, and messaging, without getting distracted by nice-to-have features early on

Ability to add and edit patient information like name, contact, age, procedure, treatment plan, and appointment details.

A task-manager-style view showing all upcoming appointments sorted by date.

Option to create reusable templates for common patient messages.

Sending reminders, caution notes, follow-ups, and medication messages directly to patients.

1. Chat Interface
A conversational flow to guide doctors through data entry and planning.

2. Treatment Plan Templates
Option to save common treatment structures as reusable templates. (Based on primary research: Many doctors follow similar steps for half their patients.)

For this project, I chose to focus only on the must-have features to keep the app lean, purposeful, and within the defined timeline.

I mapped out the key categories of information required to ensure the system runs smoothly and supports all critical workflows.

After the initial consultation, the doctor creates a treatment plan based on the patient’s condition. A treatment plan consists of multiple steps, each represented by a scheduled appointment with the doctor.

Each appointment can include the following details:

When the appointment is scheduled to take place.

A message sent before or after the appointment—this could be a reminder, follow-up, or caution note.

An internal reminder for the doctor, such as checking lab results before the appointment.

The specific name of the treatment step (e.g., “X-ray Review”).

The overall name of the treatment plan (e.g., “Root Canal”).

Additional context or instructions related to the appointment

Initially, I planned to collect scheduling details through a structured form divided into three sections:

Name, age, gender, ailment, and contact number.

Number of steps, reminders, appointment dates and times, and optional notes.

Messages to be sent before or after each step—appointment, follow-up, and caution messages

After receiving feedback, I realized that showing all fields at once could overwhelm users. To address this, I shifted to a chat-based input flow, where details are collected gradually, with contextual tables appearing for each category. For editing, the interface shows all filled-in information upfront, allowing for quick review and changes.

As outlined in the previous section on flows, this is the approach I believe can effectively address the problem statement. Below, I’ll go into more detail about specific screens and their purpose.

NetworkOcean

Saurav Chhangani