Make.com vs n8n AI Agents: Which One Wins?

Key Concepts

• AI Agents: Software entities designed to perform tasks autonomously, often leveraging large language models (LLMs).
• Make.com: An integration platform known for its user-friendly interface and extensive app integrations.
• NADN (presumably n8n): An integration platform known for its flexibility, advanced features, and focus on developers.
• RAG (Retrieval-Augmented Generation) Systems: AI systems that combine information retrieval from a knowledge base with generative language models to provide more accurate and context-aware responses.
• Vector Databases: Databases optimized for storing and querying vector embeddings, which are numerical representations of data used in machine learning.
• Pinecone: A popular vector database service.
• Multi-Agent Systems: Systems that involve multiple AI agents working together to achieve a common goal.
• MCP (Model Context Protocol): A protocol that allows AI agents to interact with software applications in a more flexible and intuitive way.
• Sub-workflows/Sub-agents: Secondary workflows or agents that the main agent calls upon to perform specific tasks.
• Global AI Agents: Reusable AI agent templates that can be used across multiple scenarios.
• Memory Stores: Components that allow AI agents to retain information and context over time.

Comparison of Make.com and NADN AI Agents

Overall Assessment

• NADN currently holds a significant advantage due to its earlier entry into the AI agent market (July of last year) and more mature feature set.
• The video focuses on the future potential of each platform and which is likely to become superior over time.
• The speaker anticipates Make.com will likely copy successful features from NADN to catch up.
• Make.com's AI agent is currently buggier, which is expected for a new release.

Target User

• Make.com: Geared towards casual users, emphasizing simplicity and ease of use.
• NADN: Aims for both ease of use and advanced functionality, targeting more sophisticated users who require greater flexibility.

Business Use Cases

• NADN excels in business use cases due to its ability to build more complex technology that Make.com currently cannot.
• RAG Systems and Multi-Agent Customer Service Builds: Highlighted as key applications for businesses.
  • These systems involve a vector database (e.g., Pinecone) that acts as a "brain" containing all relevant company information (SOPs, contracts, emails, etc.).
  • When a customer asks a question, the AI agent queries the vector database to retrieve the most relevant answer.
  • Example: A landscaping company uses the system to provide accurate pricing for lawn mowing services based on historical data.
  • Multi-agent systems can automate tasks such as sending invoices.
• Make.com currently lacks native vector database integration, limiting its capabilities in these areas.

Core Differences: Green, Red, and Yellow Categories

The speaker categorizes the differences between the platforms into three categories:

• Green (Clear Winner): NADN
• Red (Clear Loser): Make.com
• Yellow (Gray Area/Likely to Converge): Both

1. Tools Attached (Green - NADN)

• NADN allows direct attachment of tools to AI agents, simplifying the creation of simple builds.
• Make.com requires the use of sub-workflows for every tool integration, adding complexity.
• The speaker believes Make.com's linear workflow architecture will prevent it from replicating NADN's flexible tool attachment system.

2. Built-in Chat (Green - NADN)

• NADN offers a built-in chat interface for interacting with AI agents, enabling rapid prototyping and testing.
• Make.com requires integration with external messaging platforms (e.g., Telegram, Facebook Messenger), adding friction to the development process.
• The speaker is skeptical about Make.com releasing a built-in chat feature due to their lack of internal triggers.

3. Time to Build (Yellow - Gray Area)

• Make.com can be slow to build due to the need to define scenario inputs and outputs, change triggers, and enter parameters for each sub-workflow.
• NADN can also be time-consuming due to a limited number of integrations (144 compared to Make.com's 2,000+) and the complexity of working with APIs.
• The speaker cites an example of spending 6 hours trying to build something out on Pinterest only to realize after 6 hours that the API only worked for certain things.
• Make.com's native integrations for common tools can save time in many cases.

4. Vector Databases (Green - NADN)

• NADN offers native integration with vector databases (e.g., Pinecone), enabling the creation of RAG systems.
• Make.com currently lacks this capability.
• The speaker expects Make.com to eventually add vector database integration.

5. MCP (Model Context Protocol) (Green - NADN)

• NADN supports MCP, which allows AI agents to interact with software applications in a more flexible and intuitive way.
• MCP enables AI agents to intelligently determine the appropriate action to take within an application based on user input.
• Example: An AI agent can create, update, or delete Google Calendar events based on natural language commands.
• The speaker believes MCP is the future of AI agent technology and expects Make.com to eventually adopt it.

6. Global AI Agents (Green - Make.com)

• Make.com allows the reuse of AI agent templates across multiple scenarios.
• NADN does not offer this feature.
• The speaker is unsure of the practical benefits of global AI agents.

7. Memory Stores (Green - NADN)

• NADN offers multiple memory store options for AI agents.
• The speaker believes this feature is primarily useful for software engineers and advanced users.

8. Defining Parameters (Green - Make.com)

• Make.com only requires defining parameters once at the sub-agent level.
• NADN requires defining parameters twice, both when passing data into a subflow and when receiving it.

9. Testing (Green - NADN)

• NADN offers pinned and mock data for testing sub-agents, streamlining the development process.
• Make.com lacks this feature, requiring more manual testing methods.

Conclusion

• NADN currently offers a more mature and feature-rich AI agent platform, particularly for business use cases involving RAG systems and complex integrations.
• Make.com excels in simplicity and ease of use, making it a good choice for casual users with basic automation needs.
• The speaker expects Make.com to close the gap with NADN over time by copying successful features and improving its platform.
• The future of AI agent technology is likely to involve greater abstraction and more intuitive interfaces, potentially leading to a point where users can simply message a bot to perform complex tasks.