The 7 Kinds of AI Agents




These 7 categories are based on what AI agents do, not just how they work. You can think of this like a stack of increasing autonomy and collaboration.


1. 🧍‍♂️ Personal Assistant Agents

  • Function: Help individuals with tasks.

  • Examples:

    • ChatGPT, Claude, or Pi as productivity buddies.

    • Apple's Siri or Google Assistant.

  • Common use: Scheduling, writing, quick answers.

  • Sales example: Email rewriter that helps you reply faster.

🔧 Tools: GPTs, Pi, Siri Shortcuts, Reclaim.ai


2. 🧑‍💼 Task Automation Agents

  • Function: Handle repetitive actions using APIs, scripts, or browser automation.

  • Examples:

    • Zapier bots

    • Make/Integromat agents

    • Relay.app agents

  • Sales example: When a form is filled out, assign a rep, send an email, update CRM.

🔧 Tools: Zapier, Make, Relay.app, Bardeen, Lindy


3. 🧠 Decision Support Agents

  • Function: Analyze data and suggest best actions.

  • Examples:

    • Copilot in Excel or Power BI

    • Salesforce Einstein suggestions

  • Sales example: "Here's the top 5 deals at risk" or "this lead is hot because…"

🔧 Tools: ChatGPT + Docs/Sheets, Tableau GPT, Salesforce GPT


4. 🤖 Autonomous Work Agents

  • Function: Complete complex workflows without needing much (or any) supervision.

  • Examples:

    • AutoGPT and BabyAGI (experimental)

    • Devin (Cognition Labs)

    • AI SDRs that do outbound

  • Sales example: Book meetings, send multi-step outbound campaigns, follow-up.

🔧 Tools: AgentHub, Superagent, Cognosys, Devin AI (coming soon)


5. 👯 Collaborative Agents

  • Function: Work in teams (with humans or other agents), passing info and tasks.

  • Examples:

    • Multi-agent setups in LangChain or CrewAI

    • Claude + GPT-4 in the same workflow

  • Sales example: One agent qualifies a lead, another writes follow-up content, another triggers a campaign.

🔧 Tools: CrewAI, LangChain agents, GPTs with shared memory


6. 🧬 Specialist Agents (Domain Experts)

  • Function: Trained on proprietary or domain-specific knowledge.

  • Examples:

    • Legal AI agents trained on contracts

    • Web3 governance bots

    • "Chet CRM System" (😉 yours)

  • Sales example: GPT trained on your ICP, pitch decks, and email tone.

🔧 Tools: Custom GPTs, RAG systems, GPTs with memory or retrieval


7. 🌐 Networked Ecosystem Agents

  • Function: Interact across platforms and contexts—part of a decentralized "mesh" of agents.

  • Examples:

    • Agents that connect wallets, CRMs, websites, and email

    • AI agents on blockchain (e.g., Fetch.ai, Autonomous Worlds)

  • Sales example: An agent that syncs CRM status with customer on-chain behavior and adapts messaging in real time.

🔧 Tools: Fetch.ai, Autonolas, Web3 agents, AI agents using WalletConnect, ChatGPT API + Webhooks





🧩 TL;DR Table

TypeRoleSales/Marketing Example
1. Personal AssistantHelps you think/do fasterEmail reply or rewrite
2. Task AutomationAutomates workflow stepsForm > CRM > Email
3. Decision SupportGives you better insightsForecast, lead scoring
4. Autonomous WorkerDoes the job for youAI SDR books calls
5. CollaborativeWorks in agent teamsOne agent qualifies, one emails
6. SpecialistTrained on your biz/dataCustom-trained sales GPT
7. NetworkedWeb-wide or on-chain meshWallet-aware customer agent


Let's talk about your options.... Obligations... And opportunities! 


David

617-331-7852  

david@davidcutler.net 

- Growth Actions: DavidCutler.net 

- Web3 Applied: TruthRefinery.com 

- Circular Partners: CircularLabs.io