Using agents to build an agent company: Joao Moura

Joal, CEO and founder of CreAI, discusses the rapid adoption and transformative potential of AI agents, highlighting that over 10 million agents have been executed with CreAI in the last month. He explains that AI agents, powered by large language models (LLMs), can autonomously perform complex tasks, reshaping traditional software development. Joal shares his journey from creating LinkedIn posts with AI to building CreAI, which now supports various business functions like marketing and lead qualification through automated agents. He introduces CreAI Plus, an enterprise offering enabling seamless integration and deployment of AI agents as APIs, and announces new features such as automated code execution and agent training.

• AI agents are autonomous systems that can make decisions and interact with their environment without human intervention.
• They are built on large language models (LLMs) like ChatGPT, which can generate content and simulate reasonable decision-making processes.
• Agents can be designed to interact with each other, creating complex systems capable of adapting to new circumstances in real-time.

"10 million 62,000 922 over 10 million and a half that's how many agents got executed with crei in the last 30 days."

• This quote highlights the rapid adoption and execution of AI agents, indicating their growing significance and utility.

"If you get them to chat themselves or to a copy of them guess what you can leave the room you have an agent basically it can take its own decisions it can use the tools it can be autonomous."

• AI agents can operate independently, making decisions and utilizing tools without needing constant human oversight.

• Traditional automation required specific instructions and could become complex with multiple variables.
• AI agents simplify automation by adapting to new inputs and conditions without needing pre-defined connections between all variables.
• This adaptability allows for the creation of automations that were previously impossible.

"We have been building automations as Engineers for decades and usually starts pretty straightforward...but then what happens when you add C and and then you get D and this things can get complex pretty quick."

• This illustrates the complexity of traditional automation and how AI agents can streamline and simplify these processes.

"With agents you don't necessarily need to do that you don't need to connect the dots you give it the options and the Agents can adapt to the circumstances that they are met and they can do that in real time."

• AI agents can autonomously adapt to changing circumstances, reducing the need for complex manual configurations.

• AI agents require various components such as caching layers, memory layers, and training mechanisms to function effectively.
• As agents interact and form groups, additional complexities arise, such as shared caching and memory systems.
• The interaction between multiple groups of agents further increases the complexity of the system.

"When you think about the anatomy of these agents and what they look like they might look pretty simple at first...but once they start to building these things in production for real you quickly realize that you got to think about well I need a caching layer I need a memory layer I need to train them."

• Building AI agents involves more than just creating a simple model; it requires a comprehensive infrastructure to support their autonomy and interactions.

"You can go one extra level and get multiple Crews to talk to each other how that goes to say that the way that we have been building software is changing a lot."

• The interaction between multiple agent groups signifies a shift in software development paradigms, emphasizing adaptability and interconnectivity.

• Traditional software development relies on strong typing and predictable inputs and outputs.
• AI agents introduce a level of unpredictability, as they can handle a variety of inputs and produce diverse outputs.
• This shift towards "fuzzy" systems challenges conventional software testing and development methodologies.

"All the software that we have view is very strong time you start with knowing exactly the inputs that are coming in...but with AI agents and any AI apps for what it's worth everything is fuzzy."

• The transition from predictable systems to "fuzzy" AI-driven systems requires a reevaluation of software development practices.

"These models are basically black box and you don't necessarily know what's coming out of it and you know what I love it."

• Embracing the unpredictability and complexity of AI agents can lead to innovative solutions and new ways of thinking about automation and software development.

• CreAI is a production-ready library designed to build and orchestrate multi-AI agent automations.
• The platform supports over 10 million agents running every month, showcasing its scalability and production readiness.
• CreAI's development provides insights into diverse use cases and applications by users.

"CreAI is a production-ready library to build and orchestrate multi-AI agents automations."

• This quote introduces CreAI as a robust library for managing multiple AI agents, emphasizing its production readiness and extensive capabilities.

• The founder, Joe, initiated CreAI following a personal journey that began in Brazil, with significant influence from his wife.
• The idea stemmed from automating LinkedIn posts using agents, which led to increased engagement and the realization of AI's potential in automation.
• The founder's background includes working at Clearbit before starting CreAI, highlighting his technical expertise.

"The way that I build this company has been a very interesting Journey; everything started back in Brazil."

• This highlights the founder's personal and geographical journey, emphasizing the origins and motivations behind CreAI's creation.

• Initial challenges included handling bugs, hallucinations, and errors during the development of CreAI.
• Despite these challenges, CreAI gained significant traction, achieving over 16,000 stars on GitHub and forming a large community on Discord and Reddit.
• The growing interest from engineers and companies necessitated scaling the platform to meet demand.

"We start to getting bugs like rabbit hole reports hallucinations two errors... but turns out with that also CreAI gained a community over 16,000 stars in GitHub."

• This illustrates the technical challenges faced during development and the subsequent community growth and support that CreAI received.

• The solution to scaling CreAI in a competitive market was to leverage AI agents themselves.
• The founder began by building a marketing crew composed of various roles, such as a content creator specialist and a chief content officer.
• This strategic approach highlights the use of AI for internal operations and scaling efforts.

"The answer was in front of me all along; we need agents."

• This conveys the realization that the same AI technologies CreAI develops could be used to address its own scaling challenges, showcasing a practical application of its technology.

• The speaker describes the process of developing a marketing team to enhance content outreach.
• The team investigates online platforms and previous experiences to create impactful content.
• The strategy led to a significant increase in viewership within a short time frame.

"I'm going to shoveling rough ideas that kind of suck and I want to get something great so they're going to check acts and check LinkedIn and what other people were talking about this."

• The speaker emphasizes the importance of refining ideas through research and collaboration to create effective marketing content.

"We got 10x more views in 16 days 60 freaking days and I was loving it."

• The strategy was successful, leading to a tenfold increase in views over a relatively short period.

• The speaker discusses the creation of a lead qualification team to better serve potential clients.
• This team analyzes lead responses, compares them with CRM data, and provides comprehensive insights.
• The approach resulted in a high volume of customer calls, which the speaker found valuable.

"I'm going to bring up a lead analyst expert I'm going to bring an industry researcher specialist in a strategic planner I'm going to wrap them together into a lead qualification crew."

• The speaker assembled a specialized team to improve lead qualification, enhancing client engagement.

"I end up doing 15 plus customer calls in 2 weeks it was crazy you know what I don't regret it I love it."

• The intensive lead qualification process led to numerous customer interactions, which the speaker views positively.

• The speaker describes expanding the team structure to include other areas like code documentation.
• The use of agents for documentation and other tasks increases efficiency and allows for further expansion.

"We have marketing we have lead qualification let's do code documentation so if you try crew all those docks we didn't write it agents do it for us."

• The speaker highlights the efficiency gained by using agents for documentation, enabling further growth.

• The speaker mentions endorsements from reputable companies and investors as a testament to their success.
• There is a strong belief in the continued growth and impact of AI agents in the future.

"These are some of the companies they're now building with crew they're using crew and it's insane to me."

• The speaker is impressed by the adoption of their methods by various companies, showcasing the success of their approach.

"Believe some of our investors darh CTO of hubs spot or Jack outman I mean they can vouch for us we're doing pretty well."

• The speaker cites endorsements from notable investors as evidence of their success.

• The speaker discusses future developments, including enabling agents to build their own tools.
• A focus on code execution and automation is highlighted as a key area of innovation.

"We are working with code execution what means that in the new version all you got to do is create an instance of automated coder command line code executor."

• The speaker is excited about new capabilities that allow agents to execute code, enhancing their functionality.

"All you got to do it's one flag allow code execution that works your agents can code now you don't got to worry about this."

• The speaker describes a simplified process for enabling agents to code, highlighting ease of use and increased capability.

• Introduction of a new CLI feature called "Train Your Crew" aimed at enhancing employee training.
• The feature allows for consistent results over time by embedding instructions into the memory of agents.
• Emphasis on the importance of consistency in performance and results.

"Why not do that with her crew so you can get consistent results over time?"

• Highlights the analogy of training a crew like training a new employee to achieve consistent outcomes.

• The platform supports a wide range of third-party agents, promoting inclusivity and flexibility.
• Users are encouraged to integrate various agents like Yama index, Link sh, and Autogen agents.
• The platform offers shared memory and tools across all integrated agents.

"We want all the agents so bring them all. We're a universal platform."

• Emphasizes the platform's goal to be inclusive and universal, allowing the integration of diverse agents.

• Announcement of a new version release that is currently live for users to try.
• Encouragement for users to test the platform and experience the new features firsthand.

"You can try it today. We just shipped the version before I come on the stage."

• Invites users to engage with the new version and explore its capabilities.

• Introduction of a 2-hour course available at learn.crew.com to educate users about Crew AI.
• The course aims to provide comprehensive knowledge and understanding of the platform.

"We put together a 2-hour course on how to learn about Crew AI."

• Offers a learning opportunity for users to deepen their understanding of the platform and its features.

• Introduction of Cre Plus, an enterprise offering for building and deploying agents efficiently.
• Allows users to create agents, push them to GitHub, and convert them into APIs within minutes.
• Emphasizes features like autoscaling, bearer token protection, and private VPC for production use.

"With Cre Plus now you build your Crews the way that they're running your terminal."

• Highlights the enterprise-level capabilities of Cre Plus, enabling efficient agent deployment.

• Special offer for the first 50 companies that sign up, providing access to Cre Plus within 24 hours.
• Introduction of a pre-built crew that users can deploy, simplifying the initial setup process.

"For the first 50 companies that signed up using this link, we're going to give you access to Cre Plus in less than 24 hours."

• Encourages prompt action from companies to take advantage of the offer for quick access and setup.