AI Tinkerers São Paulo - Google Deepmind Hackathon

Team

NODE

This team is at maximum capacity.

Project Concept

NODE (Civil Survival Network)

As of March 2026, more than 120 million people are forcibly displaced worldwide due to armed conflicts, political instability, and climate-related crises. In active conflict zones such as Gaza, Ukraine, and Sudan, civilians face a critical lack of real-time and reliable information, along with the collapse of communication infrastructure.

Access to essential resources becomes unpredictable, and even short-distance movement can represent a life-threatening decision.

Limited access to water, food, shelter, and medical aid
Disrupted or nonexistent internet and telecom infrastructure
High uncertainty when navigating unfamiliar or unsafe areas
Lack of trustworthy, real-time local intelligence

Most existing solutions depend on centralized systems that fail precisely when they are needed most.

NODE addresses this gap.

NODE is a decentralized, edge-powered survival network designed to operate in infrastructure-constrained environments. It is an offline-first platform that enables civilians to discover nearby resources, make safer navigation decisions, and coordinate locally without relying on internet connectivity.

The system is built on three core capabilities:

On-device AI that processes and interprets data directly at the edge
Peer-to-peer communication through mesh networking using Bluetooth and Wi-Fi Direct
A dynamic, crowdsourced intelligence layer that updates in real time

Users can contribute updates through text, voice, or images. These inputs are processed locally by lightweight AI models that classify safe and unsafe areas and prioritize information based on recency, frequency, and cross-validation with nearby users.

This creates a continuously evolving map of reality, even in disconnected environments.

NODE integrates offline geospatial awareness through locally stored maps enriched with real-time overlays. These overlays highlight resource clusters and risk zones, allowing the system to recommend safer routes rather than simply faster ones.

The goal is to transform fragmented local signals into actionable intelligence that supports decision-making under uncertainty.

After all, when infrastructure fails, the network should not disappear. It should re-emerge through the people who depend on it.

Entry

Status: Submitted

Last saved: April 12 at 5:00 PM -03

View Entry

Team Roster (team is at max capacity)

You must be registered for the event to view the team message board.

Davi Nascimento de Jesus Team Lead RSVP Approved

Sales Productivity Intern at Ambev

Aqui está uma versão em inglês, utilizando um vocabulário corporativo forte e fluido, ideal para o seu LinkedIn ou currículo: As Product Strategy and Design Lead, I orchestrated the project’s strategic roadmap while spearheading the visual communication and presentation design. I was responsible for bridging the gap between business goals and user-centric design, ultimately architecting the storytelling and delivering high-impact pitches that translated complex technical concepts into a compelling vision for stakeholders and investors.

Davi Jesus is a Computer Engineering student at the Institute of Technology and Leadership (Inteli) in Brazil and a data innovator focused on transforming complex datasets into strategic decisions. His work sits at the intersection of AI, product strategy, and communication. He has developed projects applying AI to climate resilience, financial systems, and public sector decision-making, including Clima.Seguro and CarbonGuard AI. His work has received 20+ awards in science and innovation, including national hackathons. Alongside his technical work, Davi focuses on translating technical systems into products that people can understand and use. He currently works with sales productivity strategy at Ambev and leads initiatives in research and innovation communities.

AI for real-world decision systems, especially in climate resilience, public sector innovation, and financial infrastructure. I’m interested in collaborating with builders working on applied AI, data platforms, and tools that make complex systems understandable to decision-makers. I’m also excited to explore emerging interfaces between AI, product design, and human-centered communication.

I’m currently experimenting with AI systems that turn complex datasets into decision tools. Recent projects include Clima.Seguro, a govtech platform that integrates public climate APIs to help Brazilian municipalities anticipate disaster risk, and CarbonGuard AI, an ML + blockchain approach to detect fraud in carbon credit markets. I’m also exploring AI-driven storytelling tools that translate data into actionable narratives for public and private decision-makers.

David Deodato Alvarenga Nascimento RSVP Approved

IA developer at LogicaAI

I played an active role in outlining the project's development roadmap, requirements engineering, product strategy, and system architecture. Working alongside the team, I contributed to a phased execution plan that perfectly aligned with our core strategic goals by turning complex user challenges into concrete technical specs and robust designs. To support this workflow, I actively relied on Google AI Studio for prototyping models, Google Colab for processing backend data, and Google NotebookLM to structurally organize our concepts.

David Deodato Alvarenga Nascimento is an IA developer at LogicaAI.

Pedro Marcos Ramos RSVP Approved

AI Engineer at Foursys

Actively participated in defining the project's product strategy, system architecture, requirements engineering, and development roadmap. Leveraged Google NotebookLM for conceptual organization, Google AI Studio for model prototyping, and Google Colab for backend execution and data processing. By collaborating closely with the team to translate complex user needs into robust architectural designs and actionable technical specifications, I contributed to mapping out a clear, phased execution plan that ensured the final solution was perfectly aligned with our strategic goals.

Backend Developer with 6 years of experience in software development, actively participating as a technical reference for the team, contributing to design and architecture decisions and ensuring quality in delivery. Experience in Java and Spring Boot for building REST APIs and microservices, with experience in system integration, CI/CD, and good software engineering practices. In recent years, I have expanded my work into the world of Artificial Intelligence, exploring Python, Node, LLMs (Large Language Models), and SLMs (Small Language Models) applied to real-world products and solutions.

Innovation, Social Media

A social network where people connect their personal hardware to form distributed AI clusters — visualized as a living forest of bioluminescent mushrooms growing on a dark planet, connected by filaments that pulse in real time. Install it. Your device becomes a mushroom in the Fungi network. Connect to clusters of people who believe in the same project. Together you run AI that none of you could achieve alone — and you see it happening in real time, like a living forest.

Leonardo José Silva RSVP Approved

AI Engineer at Itaú Unibanco

I contributed to the team as a Software Engineer focused on AI engineering. I was involved in developing the solution from the initial code to local testing, and later implemented an online MVP so others could use it. I also provided technical insights, evaluated the solution’s practicality, and helped assess its feasibility.

Specialist in the end-to-end development, architecture, and research of solutions in Generative Artificial Intelligence, including AI agents using CrewAI, LangGraph, and LangChain, MCP servers, computer vision, NLP (Natural Language Processing), predictive Machine Learning models, and prompt engineering. In recent months, I have worked on building APIs and automations with generative AI, designing agent workflows, developing multi-agent networks, creating chats with RAG capabilities, and solving

Multi-agent conversational architectures, agent orchestration (ADK, LangChain, LangGraph, CrewAI), RAG-enabled APIs, Model Context Protocol (MCP) servers, production ML systems, computer vision, NLP, predictive machine learning, prompt engineering, microservices deployment, cloud-based AI solutions (AWS/Azure), SQL integration for agents.

Current projects include building multi-agent conversational architectures and RAG-enabled APIs at Itaú Unibanco using LangChain, LangGraph, and CrewAI. Key technical work includes **ProductAI**, **AgentSight**, and **LangGraph-agent-SQL**, focusing on agent orchestration and SQL integration. He is also developing **MCP-Guide** for Model Context Protocol servers and **RAG-new-algorithm**, utilizing a stack of Python, FastAPI, Docker, Kubernetes, and Azure OpenAI for production ML systems.