A new generation of crypto-ecological AI created by EVG: Edgen AI

Author: Gunkan Gunkan@messari.io · Research Analyst

Report summary

• Edgen is a native AI -driven market intelligence platform designed to help traders gain institutional-level insights and seize high-value market opportunities. Edgen integrates blockchain data, social sentiment and real-time insights.
• Currently, the platform's tools include Edgen Search, Radar , and Insights . Each module studies emerging narrative trends, monitors token trends, and produces high-value market insights.
• EDGM is Edgen 's proprietary multi-agent orchestration system. EDGM dynamically routes users to search the most relevant data streams, analyzes agents and large language models (LLMs), and indexes the market, thus providing actionable market analysis.
• “Aura Farming” is an incentive system that rewards users who generate and disseminate high-quality insights. Edgen gives “Aura” points based on user contributions to system improvements and provides users with a transparent analyst rating to prove their “Aura” levels to others. These contributions help improve Edgen’s orchestration layer and enhance the platform’s ability to prioritize, route and contextualize information required by users.

Introduction

Everest Ventures Group (EVG) is a venture capital studio and technology company focused on building digital infrastructure in the cryptocurrency and artificial intelligence sectors. The company is headquartered in Hong Kong and has incubated more than a dozen projects, including Aspen DigitaI, T-Rex, Mugen,

LiveArt and most recently Edgen AI. The company was co-founded by AIIen Ng and Ruby Cheng, who have more than seven years of experience in EVG, focusing on building scalable, crypto-native applications for retail and institutional markets. EVG is committed to becoming a long-term builder of the basic platform to achieve new forms of ownership, coordination mechanisms and intelligent paradigms.

Edgen is a project under EVG, committed to solving the problem of fragmentation of encrypted market information and assisting retail investors in obtaining real-time and institutional-level market intelligence. The platform helps users analyze emerging trends and asset dynamics and discover executable investment insights. As an information orchestration layer and application interface, Edgen integrates large language models (LLM), agents (agents) and real-time data flows to provide users with context-friendly real-time information.

Edgen AI

In today's crypto market, massive data is scattered across multiple sources, making it difficult to form smart decisions. Key information driving the market, such as token issuance, protocol upgrades, liquidity changes and narrative trends, is scattered on different platforms such as on-chain data, social media and dashboards. Although these data are publicly available, they are highly fragmented and lack of integration, which undoubtedly reduces user analysis speed and decision-making efficiency.

Edgen solves the problem of data fragmentation through the synergy between three core dimensions: on-chain activities, social consensus and artificial intelligence-driven analysis. At present, these three major fields not only evolve at a high speed, but also influence each other in a nonlinear way. Edgen calls this complex interaction the “Three-Body Problem”, believing that ignoring any of these dimensions will create a critical blind spot for traders and researchers.

To solve this problem, Edgen developed the Intelligent Decision Map (EDGM). As an AI orchestration system, each user's search is coordinated through modular tools, intelligence and carefully planned data flows. EDGM transforms fragmented workflows into dynamic, context-aware processes that provide timely and ground-based insights based on market conditions and user behavior. The model is also constantly optimizing the prompt structure, analysis results and performance.

The platform's current capabilities include Search (Instant Search), Radar (Token Radar) and Insights (Group Insights), serving as a wider array of smart entry points. Each tool is committed to solving a key problem in market signal flow:

• Instant search : Search for cross-chain data, news and social sentiment
• Token Radar : Monitor market dynamics and identify early financial and social signals
• Group Insight : Collecting predictions from top analysts and proven by humans and AI

Aura __Farming provides the core driving force for Edgen's collective AI layer. Users contribute to the system by training models and disseminating high-quality market insights, and obtain "Aura II points as quantitative proof of influence. These contributions will be used to continuously optimize Edgen's auxiliary models, improve prediction accuracy, and build a decentralized historical record to transparently trace the content contribution and time signals of each participant. Over time, this mechanism ultimately forms a positive cycle of collective intelligence "more smarter than ever: human users train AI, and AI provides more accurate signals, thereby improving the intelligence level of the entire ecosystem.

Technical Architecture

The core of the Edgen architecture adopts the EDGM model, a multi-agent orchestration system, which transforms the platform from a static research interface to a dynamic, modular and composable intelligent engine. Unlike relying on a single large language model to generate results, EDGM handles user searches through intelligent coordination tools, agents, workflows, data flows and large language models to provide institutional-level market intelligence in real time. Edgen's modular design is designed to ensure that it can adapt the best model to the search or task, integrate algorithmic breakthroughs in new models and modules on a real-time basis without retraining or major system transformations.

As a lightweight inference engine, EDGM can intelligently detect user intent, automatically match the most suitable AI assistant and model protocol tools, and integrate the most relevant on-chain and social data. It optimizes the output results in three dimensions: reply speed, accuracy and information relevance. Users only need to enter the search intention (such as "Analyze BTC technical trend II or "Analyze the reasons for the current rise of ETH? II), and EDGM will automatically analyze and clarify user needs, dynamically optimize the processing process through the agent system, and continuously improve itself.

Unlike traditional models that require frequent retraining, EDGM does not modify or fine-tune the underlying large language model. Instead, its focus is on improving input, optimizing search structures, and selectively using analytical tools. Therefore, Edgen can achieve efficient scaling and automatically benefit from the increasing capabilities of the underlying large language models when new models and modules are added to the platform.

EDGM multi-layer architecture background:

• Cognitive layer : Hosting the core orchestration logic of EDGM, efficient routing search using decentralized computing power network (DeGPU) through Edgen's tools and proprietary data.
• Agent layer : Drives AI assistants focused on specific tasks such as macroeconomic trend analysis, wallet behavior monitoring, social sentiment tracking or meme coin tracking.
• Execution layer : Directly drive operations based on the analysis results of EDGM. The core functions include portfolio rebalancing, strategic deployment and automatic trading execution.
• Data layer : Generate customized analysis results based on user-specific inputs (such as portfolio composition, trading history, and social activity patterns) to provide highly relevant insights.

When processing each search, EDGM will dynamically activate the corresponding system layer by intelligently analyzing instruction intent and complexity. Its breakthrough lies in the use of non-fixed paths and dynamic routing searches to generate relevant results.

Edgen Instant Search

Edgen Instant Search is a discovery engine where users can explore token dynamics, market narratives and track analysis through instructions. It can understand context and user intentions and filter out high-value results from the entire encryption field. With the support of EDGM, Edgen Instant Search, as a common intelligence layer, analyzes problems through Edgen's multi-agent strategy to provide fast and relevant insights.

Edgen Token Radar

Edgen Token Radar is a real-time monitoring tool for tracking token behavior changes, emerging narratives, and patterns across ecology and asset classes. By continuously obtaining market data, the token radar will show assets that perform well or deviate from conventional activities.

Edgen Group Insights

Edgen Group Insight is a selected level of market comments that integrate algorithms to filter content with analysts’ opinions. This module aggregates high-value information posts, charts and investment arguments, prioritizes content relevance, interactive data and user-defined interests, helps users interpret the original market signals, and filters accurately reflect the current market momentum and emotional changes.

Group insight content is published through social channels and is initially filtered by Edgen AI. The selected content is further verified by the "Aura Farmingˮ system. High-value contributors have gained exposure and reputation improvements due to their continuous output of high-quality predictions.

Aura Farming

Aura __Farming is an incentive system designed by Edgen to coordinate community participation and help the development of platform AI. Its goal is clear. On the one hand, users need to provide high-quality insights, and on the other hand, these contributions are converted into data and continuously optimize Edgen's intelligent layer.

Aura points are non-transferable reputation scores that quantify users’ contributions to the Edgen AI system. Users earn points by submitting valuable market insights or disseminating Edgen-related content. These points include both Proof of Training and Proof of Distribution

The (Proof of Distribution) function jointly builds the basic mechanism of Aura Farming.

Users can obtain Aura points through the following two activities:

1. Training AI: Release original, high-value market insights.
2. Distribution content: spread insights on external platforms and improve communication and participation.

This dual-track design cleverly converts user participation into a feedback mechanism, which can not only strengthen model performance but also expand the platform's influence. As users contribute content or promote Edgen, the system can learn, adapt and expand the data set. In the long run, the process will form a self-reinforcement closed loop, and the collective wisdom of the user group directly shapes the output of AI.

Aura Distribution Framework

Aura points are derived from a fixed prize pool, awarded to users who meet clear contribution criteria. The total prize pool includes two categories: distribution and training. Under each category, content that meets the criteria must undergo multiple rounds of AI and manual review to determine the points reward amount. This process is designed to ensure that only trustworthy, high-quality contributions are commended. The qualifications for obtaining points are as follows:

• Distribution Pool (30%): Reward users who successfully share influential encryption insights through Twitter/X. The user needs to first

Twitter/X profile is bound to the Edgen platform. The post then must mention crypto tokens, @EdgenTech official account, filtering through AI relevance filters, and reaching specific impressions within 24 hours. This mechanism encourages effective information dissemination and helps Edgen improve native visibility.

• Training Pool (70%): Reward original, high-value insight contributions to improve Edgen's AI model. The evaluation is carried out in three layers:

1. AI Rating: Automated detection verifies originality, token relevance and writing quality. Only when the minimum threshold is reached can you enter the follow-up review.
2. Community Rating: Content that meets the criteria is scored by high Aura users and reputable community auditors. Normalization algorithms can guarantee the fairness and consistency of voting.
3. Expert Assessment: Excellent expert users perform final screening, review and optimize top-notch high-quality content to ensure high standards of quality. Although this link is important, its weight is lower than the community score.

Roadmap and strategic outlook

Edgen AI is in an active development phase, planning several features designed to expand the practicality of the platform. The official timetable is about to be announced, and its preliminary AI assistant roadmap includes:

• MacroDecoder: provides concise macroeconomic dynamics, interpreting crypto market activities in a broader context of global trends.
• Stocks AnaIyzer: Refining feasible insights in traditional financial markets and bridging cross-market perceptions in the crypto and stock sectors.
• Onchain Sniper: Monitor wallet activities and token flows in real time, improve transparency, and assist tactical transaction decisions.
• Technical Analysis: Provides structural technical analysis signals, assisted pattern recognition and momentum tracking.
• Ecosystem Whisperer: Automate summary of on-chain activities, simplify research processes, and accelerate user workflow.
• TwitterGhostwriter: Featured real-time social comments to capture the narrative and emotional turn that emerge across platforms.
• BS Generator: Balances signal value with community content, reflecting cultural and informal elements in crypto discussions.
• TGEDector (TGE detector): discovers early projects and pre-issued tokens to help users seize the initiative before accepting the broad market.

In addition to a single function, Edgen’s roadmap is led by three long-term strategic principles : progressive decentralization, modular composability, and the transformation from a closed SaaS model to an open, license-free intelligent layer. This strategic direction is reflected in the core infrastructure projects in two major developments:

Edgentic __Markets and Edgen __Application chain .

Edgentic Marketplace

Edgentic Marketplace’s vision is to build an open ecosystem that carries AI-powered Co-Pilots and transaction agents. Allow third-party developers and analysts to develop and customize modular agents based on EDGM systems and achieve profitability. These AI agents cover automated execution robots, airdrop farming tools, or advanced analytics modules that integrate on-chain with social signals. The Marketplace will provide

SDK, enabling seamless integration with external applications and decentralized front-ends (DeFi Frontends). This program supports a wider ecosystem composability and creates new channels for the community through Edgen’s Agentic Stack.

Edgen AppChain

Edgen AppChain is a blockchain designed to support governance, reputation tracking and composable infrastructure. It will record major user contributions and identity-related interactions (such as reputation changes), and is expected to support transaction execution and strategy coordination in future iterations. Operate through independent chains,

Edgen can implement tailored protocol logic to meet specific needs, such as access to Edgen's AI insights, on-chain governance modules, decentralized execution, policy coordination, and progressively implemented decentralized computing resources.

Summarize

Edgen AI provides a systematic solution to the core problem of fragmentation and overloading of encrypted market information at the technical, social and narrative levels. By integrating on-chain analysis, sentiment interpretation and AI into a unified platform, users have brought more comprehensive and flexible tools to cope with complex market environments. The modular design of Edgen AI effectively reduces information asymmetry, allowing market insights mainly targeted to institutions or industry insiders in the past to benefit ordinary users more widely.

Currently launched features (including instant search, token radar and group insights) build an adaptive intelligent workflow covering from early detection to information

The whole process of interpretation of numbers. Each module helps to reduce interference information in the market and accurately present timely and effective insights. Meanwhile, Aura

The introduction of Farming has enhanced the user's sense of participation, closely combined user motivation with platform performance, and continuously optimized the AI ​​output effect through the contribution of the community.

Edgen AI focuses the core of the architecture on data integration, user feedback and result interpretability, providing a practical and feasible model that allows more users to gain high-value market insights more easily. Against the backdrop of extremely uneven distribution of market information advantages, Edgen AI is committed to bridging the information gap so that users can make decisions more efficient and confidently.