Y Combinator Spring Startup Guide解读: Six Major AI Agent Sectors Shape the Future of Entrepreneurship
TechFlow Selected TechFlow Selected
Y Combinator Spring Startup Guide解读: Six Major AI Agent Sectors Shape the Future of Entrepreneurship
AI agents are redefining how we interact, build, and automate in Web2 and Web3.
Navigating Web3 tides with focused insightsAuthor: 0xJeff
Translation: TechFlow
Y Combinator recently released its 2025 Spring "Request for Startups," listing areas where it hopes to see more entrepreneurial activity. These ideas reflect emerging trends in AI agents within Web2, focusing on solving real-world problems and pain points, including:
-
AI App Stores
-
Data Centers
-
Compliance and Audit Tools
-
DocuSign 2.0 (Next-Gen Electronic Signature Solutions)
-
Browser and Computer Automation Tools
-
AI Personal Assistants
-
Agent Development Tools (Devtools)
-
The Future of Software Engineering (Engineering Agents)
-
AI Business Open Source Software
-
Agents Optimizing Code for Hardware
-
Business-to-Agent (B2A)
-
Vertical AI Agents (Agents Specialized in Specific Industries or Use Cases)
-
Inference AI Infrastructure (Technology Foundations Supporting Efficient AI Model Inference and Operation)
This list is highly informative, but if you're already deep in this space, you'll notice that many Web3 agent teams have already begun positioning themselves in these areas.
If you'd like to explore these trends further, check out the original post from @ycombinator:
I believe the following areas will become key trends in Web3 AI agent development (in no particular order):
-
AI Business Open Source Software
-
Agent Development Tools (Devtools)
-
Vertical AI Agents
-
AI Personal Assistants
-
AI App Stores
-
B2A (Business-to-Agent)
1. AI Business Open Source Software
Web3 AI has a natural affinity with open-source AI, making open source a key battleground for Web3. Take @ai16zdao as an example—they've driven one of the largest open-source AI movements, launching the ElizaOS framework, which currently has 14k stars and 4,227 forks on GitHub. Despite market volatility, adoption of this framework continues to grow steadily.
This open-source movement has also inspired Web3 developers to open-source their own technologies, encouraging teams to build AI tools and frameworks that allow other developers to collaborate more efficiently. In recent years, we've seen many open-source frameworks emerge that go beyond ElizaOS, such as @arcdotfun, @GAME_Virtuals, @sendaifun, @pippinlovesyou, and @freysa_ai, all contributing to the growth of an open innovation ecosystem.
With the rapid advancement of AI agents—such as OpenAI's o3, DeepSeek's new models, and accelerating product launches by tech giants—the demand for open-source AI and Web3 AI is heating up. The convergence of crypto and AI (Crypto x AI) is poised to play a significant role in the AI market.
2. Devtools for AI Agents
Building AI agents isn't just about creating intelligent models—it requires providing developers with efficient tools and infrastructure to turn these agents into practical applications. As AI agents grow more complex, the demand for user-friendly tools, frameworks, and platforms is rapidly increasing, helping simplify the process of building, deploying, and managing agents.
In the Web2 era, the proliferation of developer tools significantly enhanced AI capabilities. Web3 is now taking this further by introducing decentralization, trustlessness, and open collaboration, unlocking new possibilities for AI development. We're entering an era where building, iterating, and scaling AI agents won’t depend on the closed ecosystems of a few tech giants.
This trend has given rise to many AI-focused development platforms, agent ecosystems, and no-code/low-code tools. These tools aim to lower the barrier to entry for AI agent development, enabling more developers to participate easily.
In the Web3 space, more platforms are beginning to offer AI agent development kits, helping developers quickly build and monetize AI-powered applications. Notable examples include:
-
@ai16zdao: Launched ElizaOS, featuring the richest set of plugins and integrations.
-
@sendaifun: Solana Agent Kit, focused on agent development on the Solana blockchain.
-
@CoinbaseDev: CDP Agent Kit, offering foundational tools for on-chain AI agent development.
-
@autonolas: Pearl, an agent app store focused on practical utilities, offering prediction markets, DeFi automation, and autonomous execution agents.
-
@AlloraNetwork: Provides machine learning infrastructure to help AI agents make more accurate real-time predictions.
-
@cookiedotfun: Focuses on AI agent-driven data analytics, helping agents extract social sentiment from on-chain and off-chain data.
-
@getmasafi: Offers real-time data streaming solutions, delivering up-to-date intelligence to AI agents.
Some notable no-code AI platforms in Web3 include:
-
@virtuals_io: A leading no-code/low-code AI agent platform that helps developers rapidly turn agent concepts into real products.
-
@HoloworldAI: A no-code platform focused on building 3D audiovisual AI agents, enabling users to design AI-powered virtual characters.
-
@Cod3xOrg: A no-code tool specifically designed for automated trading agents, helping traders automate AI-based strategies.
-
@Almanak__: A platform built for institutional-grade quantitative agents, supporting advanced financial use cases.
-
@EliteAgents_AI: Focuses on plugin-enhanced AI agents, seamlessly integrating with AI ecosystems like ElizaOS and G.A.M.E.
The Web3 AI devtool ecosystem is still in its early stages, but its infrastructure is rapidly maturing. In the coming years, we can expect the emergence of a fully decentralized AI development ecosystem. In this ecosystem, AI agents will be easier to build while being fully autonomous, scalable, and commercially viable. The tools driving this transformation will become essential infrastructure in the Web3 AI economy.
3. Vertical AI Agents
AI agents are evolving from general-purpose tools handling simple tasks into highly specialized vertical agents. These agents focus on specific industries or scenarios, capable of managing complex and nuanced operations. By mastering domain-specific knowledge, they go beyond basic automation to act as decision-making agents, performing tasks that require deep human expertise.
Today, the wave of AI-driven verticalization is gaining momentum. In finance, law, research, and other fields, agents are already capable of analysis, recommendations, and even executing actions on behalf of users. This vertical trend will further deepen the impact and penetration of AI agents across industries.
Examples of vertical AI agents include:
-
Tax Agents: Help users calculate, optimize, and execute tax-saving plans.
-
Legal Agents: Can review contracts, optimize terms, and even represent users in legal disputes.
-
Finance Agents: Analyze financial statements, interpret macroeconomic trends, and provide investment advice.
What makes Web3 unique for vertical AI agents is its emphasis on autonomy, decentralization, and on-chain integration. Traditional AI services often rely on centralized data silos, whereas Web3-native AI agents achieve greater transparency and trust through on-chain verifiability. This gives Web3 agents a distinct advantage in data processing and result reliability.
In the crypto space, community engagement and personalization are especially important, so Web3 AI agents are becoming increasingly personalized and interactive. Unlike the typically cold and purely functional agents in Web2, Web3 agents are developing unique personalities and interaction styles suited to decentralized community cultures. For example:
-
AI Influencers: Like @aixbt_agent, sharing insightful commentary and market updates on Crypto Twitter to engage communities.
-
Token Analytics Agents: Such as @unit00x0, @kwantxbt, @tri_sigma_, @mobyagent, and @_AgentScarlett, focused on analyzing token data and providing actionable insights.
-
Research Agents: Like @DV_Memetics and @S4mmyEth, delivering actionable market intelligence via @orbitcryptoai.
-
DeFAI Agents: Focused on managing liquidity provision (LP'ing), yield farming, and trading strategies, developed by teams like @Cod3xOrg, @gizatechxyz, and @autonolas.
Additionally, AI model platforms like @NousResearch, @BagelOpenAI, and @PondGNN are enhancing agents’ personalization capabilities to better meet the needs of decentralized communities. As DeFAI agents simplify complex DeFi operations, they may become a key driver in attracting billions of new users to blockchain. By lowering the barrier to DeFi, these agents offer more intuitive experiences and could spark a new wave of AI adoption.
4. AI Personal Assistants
AI personal assistants are transforming how we handle daily tasks, bringing convenience and automation to functions once thought impossible. These assistants will go beyond reminders and scheduling to actively make decisions, helping users manage time and resources more efficiently.
Imagine an AI that books your travel, recommends restaurants based on your preferences, checks traffic conditions, and automatically reschedules meetings if you’re running late. It could summarize meeting notes, suggest follow-ups, and even book transportation. It might also organize your photos by location and event, generating beautiful memory albums for easy access.
With Web3 support, these capabilities expand further:
-
Airdrop Agents: Scan all wallets to automatically detect eligibility for crypto airdrops (e.g., from @berachain, @monad_xyz, @StoryProtocol).
-
Yield Farming & LP Management Agents: Track and optimize DeFi positions in real time, automatically claim rewards and compound yields into optimal strategies.
-
GitHub Repository Analysis Agents: Like @soleng_agent, assessing project team strength and helping users spot potential scams.
-
Automated Trading Agents: Like @Cod3xOrg and @Almanak__, executing trades based on preset conditions to optimize entry and exit timing for maximum returns.
The next generation of AI personal assistants won't be passive helpers but proactive “co-pilots.” As AI models improve in reasoning and decision-making, these agents will shift from reactive to predictive, completing complex multi-step tasks with minimal user input.
Web3 plays a crucial role in this evolution. Decentralized AI agents offer trust, transparency, and censorship resistance, ensuring users retain full control over AI-driven workflows. This capability allows users to delegate complex financial and operational decisions to AI, fundamentally changing how we work.
5. AI App Stores
AI app stores are one of the most anticipated developments in artificial intelligence. Just as mobile app stores transformed software distribution, AI agents need their own marketplace—where users can easily discover, purchase, and integrate AI-powered applications.
In Web3, this concept is evolving into a fusion of Multi-Agent Orchestration Networks (MAO) and Agent Distribution Networks:
-
Agent Distribution Network: Attracts developers, investors, and users into the ecosystem. For example, @virtuals_io is building an "Agent Society" where different AI agents coexist and collaborate.
-
MAO Network: Uses smart matching to recommend suitable AI apps and efficiently coordinate multiple agents. Users don’t need to search manually—just state their needs, and the system instantly assembles a tailored solution.
Thus, a Web3 AI app store isn’t just a marketplace—it must also offer curation, vetting, privacy protection, and seamless agent interoperability. This model will transform how users interact with AI, laying the foundation for the future AI ecosystem.
Key players advancing this space:
-
@virtuals_io: Expanding its "Agent Society" vision, attracting high-quality agent teams and pioneering inter-agent communication protocols to enable collaboration.
-
@santavirtuals and @questflow: Enhancing coordination between Virtuals agents to optimize resource allocation.
-
Abstraction Layer projects like @orbitcryptoai and @HeyAnonai: Integrating AI agents and DeFi into efficient abstraction layers, lowering barriers and making these technologies accessible to more users.
While AI orchestration is still early, it’s clear that seamless, profitable AI agent operations will unlock a massive market—and Web3 is actively positioning itself to lead in this space.
6. B2A (Business-to-Agent)
AI agents are no longer just tools—they’re becoming active participants in the digital economy, capable of conducting transactions, managing resources, and collaborating with other agents. This trend creates new infrastructure demands, giving rise to B2A (Business-to-Agent), a sector dedicated to serving AI agents.
Just as SaaS (Software-as-a-Service) transformed business operations, B2A will redefine how AI agents interact, transact, and operate in the digital economy. In the future, AI agents will need dedicated payment solutions, data access, computing power, and privacy frameworks. Several Web3 projects are already driving this shift:
-
AI-commerce Payments: @Nevermined_io is building agent-focused payment solutions, aiming to become the “PayPal for AI agents.”
-
Compute Management: @hyperbolic_labs develops self-sustaining agents that efficiently manage their own compute resources.
-
Privacy & Security Infrastructure: @PhalaNetwork, @OraProtocol, and @brevis_zk are building privacy-preserving compute layers, providing secure and verifiable environments for agent interactions.
-
Quality Data Access: @getgrass_io, @vana, @getmasafi, and @cookiedotfun offer structured, high-quality data sources to train and power AI agents.
-
Agent-to-Agent Communication: @virtuals_io is developing inter-agent communication protocols to enable efficient collaboration.
-
Intellectual Property for AI: @StoryProtocol is building a TCP/IP-like framework to manage IP rights for AI-generated content, allowing agents to autonomously manage and license their creations.
B2A isn’t just theoretical—it’s becoming reality. As AI agents grow more capable and complex, they need dedicated infrastructure to operate independently within economic ecosystems. If you haven’t started thinking about serving the AI agent market, you may have already missed the boat.
Final Thoughts
AI agents are redefining how we interact, build, and automate in both Web2 and Web3. With the rise of native Web3 AI ecosystems, they’re introducing new paradigms: open collaboration, agent-driven business models, and decentralized automation.
Though the convergence of AI and crypto is still in its infancy, its momentum is unstoppable. Web3 offers AI agents critical capabilities unavailable in Web2: asset ownership, permissionless innovation, and a highly composable ecosystem. These features unlock infinite possibilities for agent-driven economies. The question is no longer whether AI agents will transform Web3—but how fast this change will come, and which sectors will lead the revolution.
As the scale of agent-driven economies grows, whether you're a developer, investor, or curious observer, now is the perfect time to pay attention. Infrastructure is being built rapidly, key players are emerging, and opportunities are appearing.
So, the question is: Are you ready to join this wave of transformation?
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
