System Requirements Document (SRD)

Project Name: topaz-ai

1. Introduction

The topaz-ai project, envisioned by Umang Suthar, aims to create an advanced AI system that mimics the human brain's subconscious thinking. The system will autonomously and persistently conduct research, gather knowledge, and explore all possible scenarios and cases, continuously expanding its knowledge base. This AI will proactively generate hypotheses, ask follow-up questions, and explore conditional branches to ensure comprehensive understanding. Designed for a broad audience, topaz-ai is not limited to any specific domain and aspires to be a universal tool for knowledge discovery and problem-solving.

This document outlines the system requirements for topaz-ai, including its functional and non-functional aspects, user personas, design concepts, and technical specifications. The system will also integrate internet research capabilities to ensure it has access to the latest and most comprehensive information.

2. System Overview

The topaz-ai system is a self-driven AI designed to emulate subconscious human thinking. Its core functionalities include:

• Autonomous research and knowledge gathering from both internal and external sources, including the internet.
• Persistent exploration of interconnected topics and scenarios.
• Proactive generation of questions and hypotheses to deepen understanding.
• Optional periodic summaries of findings for user review.
• A user-controlled mechanism to pause, stop, or redirect the AI's research.
• Recursive exploration of "what if" scenarios and edge cases to ensure comprehensive coverage.

The system will leverage a custom graph-based knowledge storage mechanism integrated with WeaviateDB to manage and retrieve interconnected semantic data efficiently. It will also utilize internet research capabilities to gather real-time information and expand its knowledge base dynamically.

3. Functional Requirements

• As a User, I should be able to assign tasks to the AI for research.
• As a User, I should be able to view the AI's progress and findings.
• As a User, I should be able to pause or stop the AI's research at any time.
• As a User, I should be able to query the AI for information from its knowledge base.
• As a User, I should be able to enable or disable periodic summaries of the AI's findings.
• As a User, I should expect the AI to autonomously generate and explore follow-up questions and scenarios.
• As a User, I should expect the AI to continue researching and expanding its knowledge base even after completing a task.
• As a User, I should expect the AI to self-direct its research when encountering new topics or gaps in its knowledge.
• As a User, I should be able to allow the AI to research from the internet to gather real-time information.
• As a User, I should be able to selectively re-run tasks or trigger the AI to regenerate specific research outputs.

4. User Personas

1. General User:

   • Description: Individuals from any background seeking information or insights.
   • Needs: Easy access to a vast, continuously updated knowledge base.
   • Goals: Obtain accurate, comprehensive answers to queries and benefit from the AI's proactive research capabilities.

2. Researcher:

   • Description: Academics, scientists, or professionals conducting in-depth research.
   • Needs: A system capable of exploring complex topics, generating hypotheses, and uncovering edge cases.
   • Goals: Leverage the AI's autonomous research to save time and gain unique insights.

3. Business User:

   • Description: Entrepreneurs or organizations seeking strategic insights or problem-solving.
   • Needs: A tool that can analyze scenarios, predict outcomes, and provide actionable intelligence.
   • Goals: Use the AI to make informed decisions and stay ahead in their domain.

4. Tech Enthusiast:

   • Description: Individuals interested in cutting-edge AI technology.
   • Needs: A system that demonstrates advanced AI capabilities and provides a hands-on experience with autonomous learning.
   • Goals: Explore the AI's potential and experiment with its features.

5. Visuals Colors and Theme

The visual theme for topaz-ai should evoke intelligence, innovation, and trust. Suggested color palette:

• Primary Color: Deep Sapphire Blue (#0F52BA) – Represents intelligence and depth.
• Secondary Color: Vibrant Emerald Green (#50C878) – Symbolizes growth and knowledge.
• Accent Color: Bright Amber Yellow (#FFBF00) – Adds energy and highlights key elements.
• Background: Soft Off-White (#F8F9FA) – Ensures readability and a clean interface.
• Text: Charcoal Gray (#333333) – Provides contrast and clarity.

The UI should be minimalistic yet futuristic, with smooth transitions and intuitive navigation.

6. Signature Design Concept

"The Infinite Knowledge Spiral"

The homepage of topaz-ai will feature an interactive 3D spiral that symbolizes the infinite growth of knowledge. Users will see a dynamic, glowing spiral composed of interconnected nodes (representing knowledge points).

• Animation: The spiral will rotate slowly, with nodes lighting up as the AI explores new topics. Clicking on a node will zoom into its details, showing related topics and insights.
• Interaction: Users can drag, zoom, and explore the spiral, making it a tactile and immersive experience.
• Transitions: Smooth animations will guide users between nodes, creating a sense of flow and continuity.
• Micro-interactions: Hovering over a node will display a brief summary, while clicking will reveal detailed insights.
• Background: A subtle gradient that shifts colors based on the time of day, creating a living, dynamic interface.

This design will make the AI's knowledge exploration visually captivating and instantly memorable.

7. Non-Functional Requirements

• The system must handle concurrent user queries without performance degradation.
• The AI should operate continuously with minimal downtime.
• Data storage and retrieval must be optimized for speed and scalability.
• The system must ensure data security and user privacy.
• The UI must be responsive and accessible across devices.
• Internet research must be conducted securely, with safeguards against misinformation and malicious content.

8. Tech Stack

• Frontend: React for web interface.
• Backend: Python with FastAPI for API development.
• Database:
  • VectorDB: WeaviateDB for semantic knowledge storage.
  • RDBMS: MySQL for structured data storage.
• AI Models:
  • GPT 5.2 for user-friendly responses.
  • Claude 4.5 Opas for academic and coding tasks.
  • Gemini 3 Pro for conversational interactions.
  • Google Nano Banana for image generation.
• AI Tools: Langchain, Litellm for LLM routing.
• Orchestration: Docker and Kubernetes for deployment and scalability.

9. Assumptions and Constraints

• The system assumes a reliable internet connection for continuous research and updates.
• Initial knowledge gathering will be broad, with no domain-specific focus.
• Users will have the option to pause or redirect the AI's research.
• The system will rely on WeaviateDB for efficient semantic data storage.
• The AI's periodic summaries will be optional and configurable.
• Internet research will be conducted using secure APIs and filtered for quality and relevance.

10. Glossary

• Subconscious AI: An AI system that autonomously and persistently conducts research and expands its knowledge base.
• WeaviateDB: A vector database designed for semantic search and knowledge graph storage.
• Hypotheses: Assumptions or questions generated by the AI to explore new topics or scenarios.
• Periodic Summaries: Optional summaries of the AI's findings, provided to the user at configurable intervals.
• VectorDB: A database optimized for storing and retrieving vectorized data, often used in AI applications.
• Selective Re-run: A feature allowing users to manually trigger the AI to regenerate specific research outputs.

This document outlines the foundation for topaz-ai, a groundbreaking AI system designed to emulate human subconscious thinking. With its autonomous research capabilities, internet integration, and user-centric design, topaz-ai has the potential to revolutionize knowledge discovery and problem-solving.