IntelliResearch
AI-Powered UX Research & Insights Platform
01
Context
AI-powered UX research analysis platform that transforms hours of interviews into strategic insights in minutes, reducing synthesis time by 80%.
02
The Challenge
"Analysis Paralysis" is the silent enemy of every product team. UX Researchers were spending more time transcribing and tagging hours of interviews than connecting dots and generating strategic insights. Qualitative richness was lost in endless spreadsheets or locked away in hour-long videos no one ever watched again. The challenge was to create a tool that didn't just transcribe, but "thought" like a senior researcher, automating synthesis without losing human empathy.
My Approach
Discovery & Inspiration
I shadowed research teams in startups and corporations. I identified that the real pain wasn't data capture, but synthesis and traceability. Visual inspiration came from high-frequency financial dashboards: information density, absolute clarity, and smart alerts, but adapted to the language of human behavior.
Experience Definition
I designed a workflow that acts as a "copilot" for the researcher, structured in three levels of depth:
- Intelligent Processing: Video/audio ingestion with real-time speaker diarization.
- Augmented Synthesis: Automatic generation of "Golden Quotes", sentiment detection, and pain point prioritization classified by severity (Critical/High/Medium).
- Structured Validation: Dynamic matrices where hypotheses are automatically cross-referenced with collected evidence, allowing assumptions to be validated or refuted in seconds.
IntelliResearch Design System
A scalable design system built on a Sunset Gradient palette, Apple HIG principles and semantic tokens for AI-powered research tools.
Color Tokens
Primary, secondary, and neutral system tokens built for contrast and consistency.
Primary brand color. CTAs, active states and primary action elements
Secondary brand color. Hover states, accents and gradient endpoints
Tertiary brand color. Soft backgrounds and visual support elements
Brand light color. High-contrast surfaces and content areas
Validated hypotheses, positive feedback, system success (Apple HIG)
Medium pain points, process warnings (Apple HIG)
Critical pain points, system errors (Apple HIG)
Contextual information, tooltips and guides (Apple HIG)
Typography Scale
Primary Typeface: Inter (Interface, analysis and long-form text)
UI Component States
Interactive primitives sharing token geometry, borders, and typographic rules.
Buttons
Form Inputs
Technical Implementation
I led the definition of a robust architecture to handle sensitive data and heavy processing:
- Secure Streaming: Encrypted protocols for transmission and storage of confidential user material.
- Advanced NLP: Integration of Large Language Models (LLMs) fine-tuned specifically for UX and Product terminology.
- Real-time Feedback: Websockets to allow other team members to add notes or marks during a live interview.
- Optimization: Lazy loading was implemented for massive transcripts and virtualized rendering for lists of thousands of data points, ensuring fluidity even in massive projects.
Interactive Prototype & UI Screens
Explore key flows and high-fidelity screens from the product architecture.
Transformative Results
Reduction in post-interview processing time
Increase in reuse of old insights thanks to a centralized semantic search engine
IntelliResearch transformed the researcher's role, freeing them from administrative tasks to empower their strategic capacity. We demonstrated that AI does not replace human intuition, but amplifies it, allowing us to listen to more users with greater depth and turn that empathy into better digital products.
Similar challenge?
Have complex software that needs clarity?
Let's talk directly—I'll tell you what makes sense to tackle first.