Monad Madness Hong Kong Winners Revealed: A Quick Look at the 5 Winning Projects
TechFlow Selected TechFlow Selected
Monad Madness Hong Kong Winners Revealed: A Quick Look at the 5 Winning Projects
Yesterday, the latest edition of the Monad ecosystem project pitch competition, Monad Madness Hong Kong, announced its list of winners. The five winning projects were Ammo, GM Network, TypeXKeyboard, Morpheus Farm, and Likwid.
Navigating Web3 tides with focused insights
Yesterday, the latest edition of the Monad ecosystem project pitch competition, Monad Madness Hong Kong, announced its list of winners. The five winning projects are Ammo, GM Network, TypeXKeyboard, Morpheus Farm, and Likwid.
The winning projects primarily focus on AI, DePin, and DeFi sectors. TechFlow brings you a quick overview of these projects.
🥇 First Place: Ammo – Multi-Agent Collaboration Infrastructure
Ammo’s full name is "Architectures for Massively Multiagent Online." While the name sounds academic, the problem it aims to solve is deeply user-centric. At its core, Ammo is an innovative multi-agent system. Unlike traditional single AI assistants, Ammo equips each user with a "User Buddy" capable of continuous evolution within a complex network of agents. These AI assistants do not operate in isolation but instead learn from, compete with, and collaborate with one another across a vast network. In simple terms, Ammo aims to make AI truly understand and serve human needs. Current AI can answer questions and write code, but often feels cold and mechanical—an efficient yet impersonal tool. What Ammo seeks to achieve is making AI more empathetic and human-aware.
From a technical perspective, Ammo’s breakthrough lies in its unique training methodology. The team has developed a dedicated reinforcement learning environment (RL GYM), where AI assistants can learn not only from human feedback but also improve themselves through interactions with other AI agents. This approach draws inspiration from OpenAI Neural MMO but takes it further by applying it to real-world use cases. For example, when a user is looking for a job, the AI assistant doesn’t just recommend positions—it analyzes the user's skills, interests, and career trajectory, while collaborating with other users’ AI assistants to understand market trends, salary benchmarks, and even potential career advancement opportunities. This multidimensional analysis leads to far more accurate and valuable recommendations.
In terms of privacy protection, Ammo adopts a decentralized architecture. Each AI agent collaborates under strict privacy safeguards, ensuring personal data is never misused. Additionally, the system includes built-in value alignment mechanisms to ensure that AI behaviors remain consistent with both individual user preferences and broader societal interests.
🥈 Second Place: GM Network – AI-Powered Consumer App Focused on Longevity and Health
GM Network is an AI-driven health management platform that integrates artificial intelligence with daily wellness routines to build a "LifeFi" ecosystem.
When users record fitness data via smartwatches or fitness bands, the platform’s AI assistant analyzes this information in real time, including heart rate fluctuations, exercise intensity, and calorie expenditure. GM Network has developed a “Proof of Health” mechanism that converts healthy behaviors into quantifiable value. For instance, running 5 kilometers might earn 50 GM points, while a high-intensity hike could yield 80 points. These scores reflect overall health status and can also be converted into token rewards on the platform.
On the hardware side, GM Network partners with major device manufacturers, supporting popular wearables such as Apple Watch and Samsung Galaxy Watch. Meanwhile, GM has launched its own GM Band—a smartband emphasizing precision and comfort, equipped with real-time AI coaching features.
🥉 Third Place: TypeX – AI-Powered Input Extension Plugin
According to available information, TypeXKeyboard is an early-stage Chrome extension powered by AI and the Monad network. It collects user typing data for AI training and aims to pioneer a new “Input-to-Earn” revenue model.
The project is still in its initial development phase. The official team has stated on X that the product will launch by the end of this month.
Learn more at the official account: https://x.com/TypeXKeyboard
👨👩👧👦 Community Favorite Award: Morpheus Protocol – Native Full-Stack DEX in the Monad Ecosystem
Morpheus is a native full-stack decentralized exchange (DEX) in the Monad ecosystem, aiming to provide comprehensive DeFi solutions including fast AMM swaps, token issuance tools, and cross-chain capabilities.
The full version of Morpheus has not yet been officially released. Users can visit the website: https://morpheus.farm/matrix, connect their X account to receive Monad testnet tokens ($MON), and earn multiplier points for the Morpheus platform by completing daily check-ins and social media tasks.
More details are expected in future official updates.
👍 Team Favorite Award: Likwid – Oracle-Free Margin Trading Protocol
If you frequently use DeFi lending platforms, you’ve likely encountered this frustration: wanting to borrow using LP tokens but having to go through cumbersome collateralization and borrowing steps. Likwid leverages Uniswap V4’s Hook mechanism to allow users to directly use assets within liquidity pools as collateral for borrowing.
Unlike traditional lending protocols, Likwid uses an innovative AMM formula "(x+x')(y+y')=k" to enable oracle-free, fully permissionless margin trading, supporting self-managed long and short positions. Through a swap-based liquidation mechanism and intelligent repayment system, it ensures more reliable price discovery and defends against potential attack risks, enabling any ERC-20 token to be borrowed on the platform.
In Likwid V2, a Dynamic Interest Rate Adjustment (DIRA) mechanism was introduced. It automatically adjusts interest rates based on market supply and demand—raising rates during periods of high borrowing demand to attract more deposits, and lowering them to stimulate borrowing when demand is low—thus achieving market self-equilibrium. On the capital management front, Likwid enables diversified revenue streams through coordinated operation of lending pools, liquidity pools, and insurance pools. Additionally, 30% of protocol revenues are distributed to governance token holders.
Summary: Clear Directions, Await Further Development
From the list of winning projects, it’s evident that AI-driven applications continue to lead the way in crypto innovation, while market demand for DeFi remains strong. Even amid bearish sentiment, solid application ideas—whether new or established—can still gain user recognition.
However, some of the winning projects are still in early stages, without concrete product launches or issued tokens yet. Readers should remain cautious and conduct thorough research when exploring these projects—DYOR.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
