Products
Reason
Better Questions, Better Answers.
Cognitive Infrastructure • LearnzyLabs
An internal tool built while developing Learnzy.
Product Link
| Status | 🟢 Active |
| Category | Cognitive Infrastructure |
| Built By | Himanshu Gupta |
| Role | Solo Builder |
| Stack | React · TypeScript · Supabase · Tailwind |
| Purpose | Prompt Engineering & Clarity |
Overview
Reason is an internal cognitive infrastructure tool built inside LearnzyLabs while building Learnzy.
While interacting with AI, I realized the bottleneck wasn’t the model's intelligence—it was the clarity of my initial thought.
I kept spending more time figuring out how to ask than actually building.
Reason acts as a cognitive intermediary before you speak to AI. It takes a vague idea, asks the right clarifying questions, and returns a psychologically engineered instruction set for the LLM.
Why I Built It
While building AI products, I found myself constantly fighting with prompts.
The difficult part wasn't understanding the AI.
The difficult part was reducing ambiguity in my own thoughts.
Better AI models weren’t giving me better answers. Better questions were.
I wanted a system that forced me to clarify my intent before asking the question.
That became Reason.
Why This Project Matters
Reason wasn't built as a standalone prompt tool.
It was built because I repeatedly encountered the same bottleneck while working with AI: garbage in, garbage out.
Rather than accepting that workflow, I built an internal tool to remove the ambiguity.
Many projects inside LearnzyLabs begin this way.
Features
Reasoning Modes
- Intent: Clarify what you really want.
- First Principles: Engineer a razor-sharp prompt from scratch.
- Inquiry: The system asks you questions before transforming.
- Story: Psychologically engineered writing tailored to a specific recipient.
Cognitive Workflow
- Interactive context gathering
- Frictionless UI
- Local history & reuse
Screenshots
How It Works
Idea
↓
Reasoning Mode
↓
Better Question
↓
Better Prompt
↓
Better Output
Architecture
React (Vite)
↓
Supabase
↓
Edge Function
↓
LLM
↓
UI Output
The application separates the mental model selection from the underlying LLM transformation, allowing new cognitive frameworks to be added seamlessly.
Tech Stack
Frontend
- React
- TypeScript
- Vite
- Tailwind CSS (shadcn-ui)
Backend
- Supabase
- Edge Functions
Design Philosophy
Think first. Prompt second.
Most prompt tools optimize for rewriting text.
Reason optimizes for clarifying thought.
The design is strictly minimal and non-flashy. It removes all distractions, adopting a clean, confident aesthetic that forces focus on the cognitive process rather than the tool itself.
Lessons Learned
The biggest bottleneck in AI isn't the model.
It's human ambiguity.
Once you realize that prompt engineering is fundamentally about reducing ambiguity, it changes how you approach AI. A good tool doesn't just improve the output; it helps you understand what you wanted to ask in the first place.
That realization has influenced how I design products beyond Reason.
Roadmap
Intelligence
- Automatic mode selection based on idea
- Context persistence across sessions
Integrations
- Browser Extension for instant access
- API for universal reasoning