Brazil's AI ecosystem: how to take part and build useful connections
Taking part in Brazil's AI ecosystem starts with choosing what you need to learn or build and finding people who work on that problem. Communities, universities, companies and gatherings offer different entry points. Value emerges when a conversation helps clarify a decision, test an idea or move a project forward.
When I organize an event, one of the things I think about is who should meet whom. I discussed this in my conversation with Lucas Abreu on Aura: observing participants and taking care with introductions is part of hosting people. A relevant connection starts with understanding where each side stands. Aura, from 07:32.
For someone just arriving, I suggest giving the same attention to your own participation: have context to share, a concrete question and a willingness to contribute.
What makes up an AI ecosystem?
An artificial intelligence ecosystem brings together people and organizations that research, develop, fund, teach and apply the technology. Those producing knowledge, those building tools and those familiar with a business problem need to be able to talk to each other.
A community can bring these participants together, while a university develops research and a company provides a real setting for application. Relationships between these environments help knowledge and needs circulate.
The mapping by OCIAAM, developed by researchers associated with USP and CIAAM, allows you to explore institutions, research areas and AI collaboration networks in Brazil. It shows one part of this ecosystem: knowledge production and its connections. For an entrepreneur, it broadens the search for people to speak with beyond business events.
Choose an entry point according to the problem
“I want to learn AI” can lead to hundreds of conversations. “I want to understand how to evaluate an assistant that consults documents” helps you find someone with relevant experience.
A practical way to guide your search is to match the question with the type of person to approach:
| What you need to understand | Where to look for a conversation |
|---|---|
| How to build and test an application | Developer groups, workshops and hackathons |
| Whether a problem needs a solution | People who perform the task and potential users |
| How to assess a specific technical question | Researchers, specialists and teams that have already faced it |
| How to bring the product to market | Founders and professionals in the sector you want to enter |
For technical learning, Google's Build with AI brings together workshops and gatherings associated with local developer communities. For research, OCIAAM's atlas helps locate institutions and topics. To meet founders, look for environments that offer time together and exchanges about the challenges of building companies. Check each initiative's purpose and schedule before taking part.
Bring enough context to receive useful help
A useful introduction can fit into four sentences: what you do, what problem you are investigating, what you have already tried and what you would like to understand. This is a preparation suggestion that you can adapt to the conversation.
Imagine a hypothetical situation: a founder wants to investigate inventory losses in small grocery stores. At a gathering, she explains that she is studying how shopkeepers plan purchases and wants to meet someone who has worked with demand forecasting. Someone familiar with a similar experience is better placed to help or make a relevant introduction.
She can also offer something: explain a process in her industry, test a demonstration or share a public reference. Business knowledge has value for people building technology. A good exchange allows both sides to learn.
Agree on a next step that makes sense
After a productive conversation, record the topic and what was agreed. If both sides are interested, a message can pick up the context:
I enjoyed our conversation about demand forecasting. You mentioned that seasonality was one of the project's challenges. I have set aside the reference I mentioned; if useful, I can also share what we are learning from conversations with shopkeepers.
This is an example message, rather than a formula for sending at scale. Only make commitments you can fulfill. One conversation may end with a useful reference; another, with a joint test. The next step should reflect people's interest and availability.
When discussing distribution in Aulão 2, I highlighted the role of community, content and connections as building similar applications became easier. Being in touch with people who can use, question and recommend a solution is part of entrepreneurship. It requires relationships and an understanding of the problem. Aulão 2, from 01:08:13.
How can you tell whether participating is helping?
After a few gatherings, observe what has changed: have you defined the problem better? Found someone with relevant experience? Received criticism that changed your test? Managed to contribute to someone else's work?
These signs help you choose where to keep showing up. Regular participation lets you resume conversations and share what happened after a suggestion. For those organizing these environments, explore the other perspective in the article on how to build a community in the innovation ecosystem.
To read more about the conditions that sustain different ecosystems, see our analysis of AI infrastructure and scale in China.
If you want to get closer to these exchanges between founders and AI projects, discover AI House Brasil, an initiative by NewHack and Entrypoint. See how its approach connects with your current situation and bring a concrete question to the next conversation.