| **Independent Researcher & Strategist | Agentic Commerce · A2A Markets · Agentic Search · AI Visibility · Decision Authority | Shanghai, China** |
I am Shen Xu (徐慎), an independent researcher, strategist, author, entrepreneur, and product builder based in Shanghai, China.
My current work increasingly centers on Agentic Commerce and Agent-to-Agent (A2A) markets: how commerce changes when AI systems move beyond information retrieval and begin participating directly in evaluation, selection, action, coordination, and transaction.
I study this transition through Agentic Commerce, A2A Market Architecture, Agentic Search Optimization (ASO), Decision Authority Economy (DAE), AI Visibility, and Entity Ranking.
My work combines research, strategic consulting, product development, and hands-on implementation. Alongside books and papers, I build open-source frameworks, interactive assessment tools, agentic-system prototypes, DeepSeek Harness plugins, AI skills, decision tools, and AI-native product experiments.
Current implementation work translating Agentic Commerce, A2A markets, Agentic Search, Decision Authority, AI Visibility, and AI-mediated selection into public research frameworks and usable assessment tools.
An open research framework for understanding how markets, merchants, platforms, channels, and commercial infrastructure change when AI agents become active participants in discovery, evaluation, coordination, negotiation, and transactions.
An open-source research framework and evaluation tool for assessing how AI agents discover, understand, verify, qualify, select, and act on brands, products, companies, services, and other entities.
A scenario-based interactive assessment for practical GEO, AI Visibility, retrieval, citation, recommendation, query fan-out, and agent-ready optimization knowledge.
My product work spans mobile applications, venture building, enterprise digital strategy, SaaS, AI-native systems, and agentic decision tools.
View the full Products, Ventures & Agentic Systems timeline →
My primary research direction is increasingly Agentic Commerce and Agent-to-Agent (A2A) markets: environments in which AI agents participate directly in discovery, evaluation, coordination, negotiation, selection, and transaction processes.
The term A2A here refers broadly to agent-to-agent economic interaction and is not limited to any single A2A communication protocol.
I am particularly interested in the emerging infrastructure connecting:
The broader progression I use is:
AI Visibility → Agentic Search → Agent Readiness → Agentic Commerce → A2A Markets
Another way to express the commercial progression is:
Visibility → Eligibility → Selection → Transaction
This connects AI Visibility and Agentic Search to the larger question of how decision authority and economic activity change when AI systems participate directly in market decisions.
A2A Market Architecture is an open research framework for studying how markets, merchants, platforms, channels, and commercial infrastructure change when AI agents become active market participants.
The framework examines six initial areas:
The central research question is:
What happens to commerce when AI agents are no longer only interfaces or assistants, but become active participants in markets?
Repository:
https://github.com/ShenXuAkaEkstasis/a2a-market-architecture
My work on Agentic Search Optimization (ASO) examines the shift from traditional search ranking toward AI systems that retrieve, understand, verify, compare, qualify, select, and potentially act on entities on behalf of users.
I use the following six-stage framework:
Discover → Understand → Verify → Qualify → Select → Act
The current research-preview implementation contains 24 assessment criteria across these six stages.
Framework page:
Agentic Search Optimization (ASO) Framework
Live research preview:
https://huggingface.co/spaces/ShenXu1013/agentic-search-optimization
GitHub:
https://github.com/ShenXuAkaEkstasis/agentic-search-optimization
Product Hunt:
https://www.producthunt.com/products/aso-framework-agentic-search?launch=aso-framework-agentic-search
I study how brands, companies, products, people, and other entities are represented, retrieved, evaluated, cited, recommended, and selected by AI systems.
AI Visibility is an important starting layer, but visibility alone does not guarantee that an entity will qualify for a task, be selected over alternatives, or become usable by an AI agent.
Decision Authority Economy (DAE) is a framework for understanding what happens when people delegate parts of search, comparison, recommendation, selection, and execution to AI systems.
The framework focuses on how the redistribution of decision authority changes influence, competition, attribution, market power, and the relationship between consumers, brands, platforms, and AI systems.
I also research how AI systems evaluate the importance and authority of entities.
My 2026 arXiv paper, Representing Entity Importance in AI Knowledge Systems: A Dual-Signal Framework of Audience Evaluation and Structural Authority, examines entity importance through audience-evaluation and structural-authority signals.
Selected implementation work includes:
A2A Market Architecture — open research framework examining how markets and commercial infrastructure change when AI agents become active market participants
https://github.com/ShenXuAkaEkstasis/a2a-market-architecture
Agentic Search Optimization Framework — open-source ASO framework and interactive evaluation tool
https://shenxuakaekstasis.github.io/ASO-Framework.html
https://huggingface.co/spaces/ShenXu1013/agentic-search-optimization
https://github.com/ShenXuAkaEkstasis/agentic-search-optimization
https://aso-framework.vercel.app/
https://www.producthunt.com/products/aso-framework-agentic-search?launch=aso-framework-agentic-search
GEO Knowledge Test — scenario-based interactive GEO and AI Visibility assessment
https://github.com/ShenXuAkaEkstasis/geo-quiz-vercel-v2
https://geo-quiz-vercel-v2.vercel.app/
A2A Decision Layer — executable prototype exploring AI-to-AI decision and transaction structures
https://github.com/ShenXuAkaEkstasis/a2a-decision-layer
AI Decision Mediation Framework — framework repository for AI-mediated decision systems
https://github.com/ShenXuAkaEkstasis/ai-decision-mediation-framework
DSH AI Shopping Assistant — DeepSeek Harness plugin for product comparison, candidate ranking, realizable-price analysis, merchant/review evidence, and source-quality checks
https://github.com/ShenXuAkaEkstasis/dsh-ai-shopping-assistant
DSH AI SaaS Deal Finder — DeepSeek Harness plugin for identifying legitimate lower-cost AI/SaaS purchase options while checking region, eligibility, payment, tax, and service-policy constraints
https://github.com/ShenXuAkaEkstasis/dsh-ai-saas-deal-finder
Tencent SkillHub portfolio — published AI skills covering shopping, SaaS deal discovery, business evaluation, fundraising, reputation strategy, creative generation, and AI-native utility workflows
DepthMotion AI — AI-powered photo-to-cinematic-3D-motion product using depth estimation and parallax rendering
https://github.com/ShenXuAkaEkstasis/DepthMotion-AI
Entity Ranking Comparison — implementation work connected to my research on entity importance and structural authority
https://github.com/ShenXuAkaEkstasis/entity-ranking-comparison
A fuller product, venture, and implementation history is available here:
Products, Ventures & Agentic Systems
GitHub profile:
https://github.com/ShenXuAkaEkstasis
Hugging Face:
https://huggingface.co/ShenXu1013
From Search to Agentic Commerce: Attribution Collapse, Agentic Search, Decision Authority, and A2A Markets
Eliva Press, 2026.
The book examines the transition from search-driven discovery toward AI-mediated selection and, eventually, agent-mediated transactions.
It connects four themes:
Amazon:
https://www.amazon.com/dp/9999353231
Publisher:
https://elivabooks.com/en/book/book-5876480398
A framework for understanding how decision-making authority shifts as AI systems take a larger role in search, recommendation, selection, and execution.
Amazon:
https://www.amazon.com/dp/B0GY65JZ3H
Agentic Search Optimization: A Practical Introduction to AI Search, Visibility, and Selection
A practical exploration of how AI systems are changing information discovery, recommendation, evaluation, and selection.
Amazon:
https://www.amazon.com/dp/B0H2Y7MYNR
A Chinese-language book on AI Visibility and how organizations can improve their representation and discoverability in AI-mediated environments.
Amazon:
https://www.amazon.com/dp/B0H4XBSYQD
Social Community Marketing in China: How Douyin and WeChat Shape China’s Internet
An earlier Chinese-language book examining social media, community marketing, digital platforms, and the evolution of user attention and commercial influence in the Chinese Internet environment.
This earlier publishing work forms part of the longer intellectual path from social distribution and platform influence toward AI Visibility, Agentic Search, Decision Authority, Agentic Commerce, and A2A markets.
Author: Shen Xu
Published: arXiv, 2026
Paper:
https://arxiv.org/abs/2607.20925
Code repository:
https://github.com/ShenXuAkaEkstasis/entity-ranking-comparison
I have more than 20 years of experience across digital strategy, technology, marketing, product development, venture building, SaaS, and China market consulting.
Across agency, internet, technology, software, consulting, and entrepreneurial work, I have worked with or supported 300+ brands and organizations.
Selected experience includes:
Shanghai Vanilla Technology / 上海香草科技 — co-founder and operating role during the early mobile-Internet period; products included Banmake and Xiangge. Xiangge-related products exceeded 5 million cumulative downloads, and the company received strategic investment from Qihoo 360.
DigitasLBi Greater China / Publicis Groupe — served as Strategy Director, connecting digital strategy, product thinking, consumer behavior, and enterprise transformation.
TUNA — founder and CEO; angel-backed entrepreneurial and commercialization experience across technology, digital products, and enterprise projects.
Yueliu / 阅流 — growth and market work inside a venture-backed technology company.
TSplus — China market marketing and growth work for a French B2B software company, including localization, search visibility, content, product positioning, and market development.
Technology-company advisory work — business model, market strategy, growth, partnership, and ecosystem development for venture-backed technology businesses.
My current work combines AI research, product experimentation, open-source development, and strategic consulting, with particular emphasis on Agentic Commerce, A2A markets, Agentic Search, AI Visibility, AI-mediated decision systems, and China market strategy.
My current work centers on a broader question:
What changes when AI systems move from helping people find information to evaluating options, making selections, taking actions, coordinating with other systems, and participating in transactions on their behalf?
I explore this through Agentic Commerce, A2A Market Architecture, Agentic Search Optimization, Decision Authority, AI Visibility, Entity Ranking, and AI-mediated commerce.
The underlying transition can be summarized as:
AI Visibility → Agentic Search → Agent Readiness → Agentic Commerce → A2A Markets
Shen Xu / 徐慎
Independent Researcher & Strategist
Agentic Commerce · A2A Markets · Agentic Search · AI Visibility · Decision Authority
Shanghai, China