Protected: World’s first AI based data scientist platform
Client
UBIX Labs
Project Type
Full Time
Industry
AI/ Data Science/ SaaS
My Role
Product Designer & Visual Lead
Tools
Figma, Adobe Creative Suite, Slack
Business Oportunity
AI was powerful — but locked behind a technical wall
The global appetite for AI and advanced analytics has never been greater. Yet for most organizations, actually deploying AI remained a slow, expensive, and specialist-dependent process. Data science teams were stretched thin serving reports instead of driving innovation. Business leaders sat on mountains of untapped data, unable to act without IT bottlenecks.
UBIX saw a clear opening: deliver a platform that lets any business — regardless of size, industry, or technical maturity — become an AI-driven organization in days, not months. The opportunity wasn’t just a better analytics tool. It was a fundamental shift in who gets to do data science and how fast value is realized across manufacturing, finance, supply chain, healthcare, and beyond.
From a design standpoint, this meant the product’s credibility lived or died on usability. A platform powerful enough to replace entire data science workflows had to be simple enough that a business analyst could open it and start building immediately. That gap — between raw capability and everyday usability — was the design challenge.
Problem Statement
How do you make enterprise AI feel effortless?
Enterprise AI platforms are built for power — and it shows. Steep learning curves, fragmented workflows, and interfaces designed for technical specialists have kept AI out of reach for the majority of business users. The more capable the platform, the harder it tends to be to use.
UBIX was different in ambition but faced the same design risk: a platform spanning generative AI, reinforcement learning, drag-and-drop model building, and custom app creation could easily overwhelm the very users it was trying to empower. Non-technical users needed to feel guided, not lost. Data scientists needed depth without friction. Executives needed to see insights, not interfaces.
The core problem: design a unified experience across a highly complex, multi-module AI platform that scales from citizen data scientist to enterprise power user — without ever requiring a single line of code, a manual, or a dedicated training session.
Redesigned the UBIX Dashboard and core navigation to create a welcoming, intuitive entry point for all user types. The new layout introduced a clean workspace hub, streamlined access to key areas like ModelSpace, DataSpace, and SolutionSpace, and surfaced the most relevant resources at a glance — reducing cognitive load and helping users move from login to insight faster than ever.
Designed the ActionSpace model training workflows — including Train Test Split & Balancing, Train Regression Model, and Train Classification Model — with a focus on guiding non-technical users through complex machine learning processes step by step. Clear visual progress indicators, real-time feedback, and structured result displays made model training feel approachable, transparent, and confidence-inspiring for users of all skill levels.
Crafted the NLP Categorization Action interface to make natural language processing accessible without a single line of code. The design surfaced key metrics like F1 Score, Accuracy, Precision, and Recall in a clean, scannable layout, while concept relationship visualizations helped users immediately understand how their data connected. Alongside this, the SolutionSpace workspace was designed to give team members a clear, organized view of all active data projects — making collaboration and insight discovery seamless.
Designed the DataSpace connection interface to simplify the process of importing data from dozens of sources — from databases to third-party platforms — into a single, clean workflow. The ModelSpace pipeline view was crafted to give data scientists and business users alike a visual, node-based understanding of their entire data flow, making complex model architectures easy to build, review, and communicate across teams.
The SolutionSpace Prediction interface was designed to be the most user-facing, high-impact screen in the platform — where business users come to explore forecasts, test scenarios, and make real decisions. The layout was crafted to surface predictions and actuals side by side with clear visual contrast, interactive sliders, and accessible data storytelling — turning raw model output into something any stakeholder could understand and act on immediately.
Design Outcome
A platform complex enough for enterprise, simple enough for everyone
UBIX is a patented no-code AI platform spanning ModelSpace, ActionSpace, SolutionSpace, and the SSApp Builder. As Product Designer and Visual Lead, the work was to translate that technical depth into a single, coherent experience — one that felt effortless for a first-time business user and powerful enough for enterprise teams running production-grade AI workflows.
The result was a unified design system that made UBIX feel as credible as it was capable — attracting enterprise partnerships and the attention of industry players who recognised the product’s potential firsthand.
Oracle explored investing in UBIX — the deal didn’t close due to their internal rules and regulations, not the product. That’s the kind of signal that speaks for itself.