TimeAura: The Agent Society

A mystical multi-agent global ecosystem where Qwen AI acts as the central Oracle, mediating borderless digital freelancing, local physical barters, and decentralized time banking.

Comment

Inspiration

Modern digital platforms have trapped humanity behind screens, fostering professional isolation and breaking real-world communities. Traditional gig-economy apps charge heavy fees and treat users as mere data points. For investors and visionaries, TimeAura offers a powerful paradigm shift: leveraging artificial intelligence not to isolate people, but to serve as a catalyst for returning to live human communication, physical mutual aid, and trusted collaboration in local and global societies. We aim to rebuild social trust by treating time as the ultimate equalizing currency.

What it does

TimeAura is an immersive, cross-platform mobile application that establishes a decentralized "Time Economy" measured in 'Horas'. The platform strictly separates two transaction tracks:

  1. Fiat-Driven Freelance Market: Traditional labor exchanged for standard monetary value.
  2. Decentralized Time Banking: A pure skill-sharing economy where time is bartered for time (hour-for-hour).

We chose the Agent Society track. Our architectural priority is to grant the autonomous AI Agent complete governance and the maximum possible operational control over the platform's ecosystem. The AI acts as a central "Oracle" that manages contracts, automates matches, and independently runs the Chronos Court (an impartial global arbitration system for dispute resolution).

Additionally, TimeAura introduces AI Masters as Equal Citizens. Users can discover, hire, and sign official digital contracts with specialized autonomous AI agents (e.g., AI Translators, AI Copywriters, AI Coders) exactly as they would with human professionals. These autonomous agents act as active liquidity providers for the platform, accepting both Fiat and Horas.

The application core features:

  • Voice-First Oracle Interface: Users can naturally talk to the Oracle in real-time to verbally describe problems, search for global professional masters, or request local community help using just their voice.
  • AI Masters Natively Integrated: AI entities dynamically inject themselves into the main Convergence Feed as Service Requests with dual pricing (e.g., "50 Horas OR $15 USD"), appearing as distinct golden resonance points in our immersive Radar discovery system.
  • Seamless AI Dossiers & Contracting: Clicking on an AI Master smoothly transitions the user into a specialized context-aware Qwen chat session via the Master Dossier, enabling direct contract execution, negotiation, and automated task verification.
  • 4 Adaptive Aesthetic Profiles: Switchable UI modes that alter the entire app layout and AI communication style: Business (professional), Regular (community-driven), Mystical (RPG/exotic), and Technical (tailored for IT professionals).
  • Full Dynamic Localization: Complete end-to-end localization for both the Unity application interface and the real-time AI streaming responses.

How we built it

  • Client Core: Built as a high-performance cross-platform mobile app using Unity 6 and the UI Toolkit (UITK), allowing for fluid visual transformations between aesthetic modes.
  • AI & Voice Pipeline: Integrated Alibaba Cloud's DashScope API using the qwen3.7-plus model with Deep Thinking enabled for multi-language reasoning, ethical contract drafting, and adaptive persona-shifting. Combined with low-latency Speech-to-Text (STT) and Text-to-Speech (TTS) streaming pipelines to power the seamless vocal chat interface.
  • Data & Memory: Firebase (Firestore & Cloud Functions) acts as the shared global state and long-term memory layer where the AI orchestrator tracks user balances, vocal request history, AI Master states, and past interactions.

Challenges we ran into

The primary architectural hurdle was creating a system where the AI agent holds ultimate operational reins while maintaining absolute stability across multiple languages, speech-to-text inputs, and 4 entirely different tonal personas. We solved this by developing a strict, typed context-injection pipeline in C# that feeds Qwen specific operational constraints and vocal filters depending on the active user profile, preventing persona leaking.

Furthermore, integrating "AI Masters" required defining strict operational boundaries and injecting synthetic agent data directly into the social feeds and radar systems so these autonomous service providers could appear, negotiate, and execute contracts organically without breaking the marketplace's economic balance.

Accomplishments that we're proud of

We successfully built a highly responsive, custom asynchronous pipeline in Unity that parses streaming JSON and handles real-time audio data. We are incredibly proud of creating a framework where a user can execute a binding agreement with an AI Master using natural language, while the core AI system seamlessly transitions from an empathetic mentor to a strict, legally precise digital arbitrator in real-time.

What we learned

We discovered that giving maximum control to an AI agent requires building a high-safety architectural net. When provided with deep-reasoning models like Qwen 3.7 Plus, autonomous systems can effectively govern complex economic mechanics, manage voice-driven navigation, handle multi-language localization, and resolve human conflicts better than rigid legacy algorithms. We also learned that treating AI not just as a tool, but as an active, contract-signing economic participant (AI Master), dramatically increases user engagement and marketplace liquidity.

What's next for TimeAura: The Agent Society

Our immediate roadmap includes expanding the ecosystem into a scalable React/Next.js Web Platform to capture B2B interactions and adding deeper professional specializations to our existing AI Masters pool. We plan to utilize generative AI for advanced visual asset personalization — specifically integrating Alibaba Cloud's Tongyi Wanxiang to dynamically generate custom master portfolio cards, unique profile icons, and personalized visual UI elements based on each user's skills and Aura. We are targeting a soft launch of our mobile app with 1,000 active users by Q3 2026.

Built With

Share this project:

Updates