AI Tinkerers São Paulo - Google Deepmind Hackathon

Team

The Sailors

This team is at maximum capacity.

Project Concept

Project: SilverEcho – The Proactive Voice Companion
The Problem
Current voice devices (Alexa/Google Home) are reactive: they wait for a command. For many seniors, the challenge is social isolation and the onset of cognitive decline. They need someone to take the initiative to talk, to listen to their stories, and to monitor their well-being without being intrusive.

The Solution (What we are building)
SilverEcho is a voice agent specialized in emotional and functional geriatrics. It operates on three pillars:

Active Listening and Affective Memory: The AI initiates conversations based on the senior’s history (“How was physiotherapy today?” or “Tell me more about that trip you mentioned yesterday?”). It uses RAG (Retrieval-Augmented Generation) to maintain a “long-term memory” of the user’s stories, combating loneliness through the validation of memories.

Cognitive and Functional Support: More than alarms, it offers context. Instead of “Medicine time,” it says: “Crizan, let’s take that little red heart pill now? I’ll stay here with you until you take it.”

Voice Pattern Analysis: Uses AI to detect subtle signs of fatigue, mental confusion, or sadness through prosody (tone of voice), alerting family or caregivers if there is a sudden change in emotional pattern.

Technical Differentiators for the Hackathon
Agent Architecture: One agent for casual conversation, another for health monitoring, and a third for interfacing with the family.

Persistent Memory: Use of vector databases so that the AI ”remembers” personal details, creating a real bond of friendship.

Voice Interface (STT/TTS): Focus on low latency so that the conversation flows as if it were with a human.

How you can contribute
We are looking for builders interested in:

LLM Memory Management: Structuring how AI stores and retrieves personal memories.

Voice UX: Designing conversation flows that don’t frustrate people with speech or hearing difficulties.

API Integration: Connecting voice with alert systems (WhatsApp/SMS) for family members.

Why is this truly “social impact”?

The world is aging, and there won’t be enough human caregivers for everyone. An AI that supports, converses, and listens restores dignity and a sense of importance to the elderly, as well as providing peace of mind for the family.

Entry

Status: Not Started

Team Roster (team is at max capacity)

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

Crizan Belem Ribeiro Team Lead RSVP Approved

Software Engineer at Vivo

Crizan Belém, whose full name is Crizan Belem Ribeiro, is currently an AI Architect at GFT Technologies and also holds the title of Software Engineer at Vivo, indicating they are employed and open to work. With approximately 10 years of experience, Crizan is focused on technical architecture. Their Tinkerer role is AI Engineer (Research & Development), currently architecting multicloud Generative AI platforms and RAG pipelines at GFT Technologies. This work involves building agentic orchestration systems using LangGraph and LangChain, integrating vector databases, and implementing graph-based solutions with AWS Neptune, alongside engineering scalable ML platforms utilizing various AWS services. Crizan studied Engenharia de Computação, Ciência da Computação (Postgraduate), Física, and Física (Doutorado) at institutions including Universidade Metodista de São Paulo, Alura, Universidade Cruzeiro do Sul, and Universidade Estadual Paulista Júlio de Mesquita Filho. They are located in São Paulo, Brazil, are open to introductions, and prefer contact via email.

Multicloud Generative AI platforms, RAG pipelines, agentic orchestration, LangGraph, LangChain, vector databases, graph-based solutions, AWS Neptune, scalable ML platforms, MLOps, Graph Neural Networks, distributed data pipelines, technical architecture, AI/ML research and development.

Currently architecting multicloud Generative AI platforms and RAG pipelines at GFT Technologies. Tactical focus includes building agentic orchestration systems using LangGraph and LangChain, integrating vector databases, and implementing graph-based solutions with AWS Neptune. Engineering scalable ML platforms and distributed data pipelines utilizing AWS services like ECS, Lambda, Glue, and Redshift to operationalize LLMs and Graph Neural Networks for enterprise applications.

ANDRE HENRIQUE CORREA RSVP Approved

Technology Specialist at CLARO

André Henrique Correa is a Software Developer focused on non-AI projects, currently serving as an Analista Sênior de Soluções e Automação de Sistemas at Aché Laboratórios Farmacêuticos, with over 15 years of experience. His work involves leading systems evolution and automation, including API development for integrations between Qualitor and Microsoft Teams, implementing Azure backup and disaster recovery, and configuring cloud security controls. He holds a degree in Sistemas de Informação from Universidade Anhanguera São Paulo. André is open to introductions and seeks partnerships, technical architecture guidance, customers, investors, and sponsorships, preferring contact via email.

Partnerships, technical architecture, customers, investors, sponsorships, systems integration, automation, cloud infrastructure (Azure and AWS), API development, cybersecurity, disaster recovery, DevOps, monitoring, and operational stability.

Currently leading systems evolution and automation at Aché Laboratórios Farmacêuticos, focusing on API development for integrations between Qualitor and Microsoft Teams. Key activities include implementing Azure backup and disaster recovery, configuring cloud security controls, and developing Power BI dashboards. Technical work involves managing systems integration, monitoring operational stability, and utilizing tools like Antares Vision and Evolutio within regulated industrial environments.

Thais Coffani Costa RSVP Approved

Analista de field marketing at Indorama Ventures

Graduated in chemical engineering and master's student in production and systems engineering Researcher in the area of AI in the field of chemical engineering and agriculture

AI for agriculture, AI to create impact at society

Articles that I have published: Predicting Biomass Gasification Activation Energy Using Artificial Neural Networks – 8th International Congress on Technology for the Environment (CITMA) Demographic Prediction Through Carbon and Nitrogen Isotope Composition of Fingernails Using Artificial Intelligence – International Forensic Science Conference (InterForensics) Applications of Artificial Intelligence in Wastewater Management in Brazil: A Systematic Review – National Meeting of Production

Rute Andrade Oliveira RSVP Approved

Estagiaria at Prefeitura Municipal de São José dos Campos

I am a humanities student with experience in social work, administration, and legal support. My career path is marked by an interest in promoting social impact, especially in access to rights and the organization of services for vulnerable populations. Currently, I work as an intern handling legal and administrative demands, developing analytical and organizational skills. I have a strong interest in technology and artificial intelligence applied to social good, and I seek to participate in initiatives that integrate innovation and impact, such as this hackathon.

A for social impact; access to justice; public policies; social inclusion; use of data for decision-making; technology applied to the public and third sectors; automation of social and legal processes.

Currently, I work as an intern focusing on legal and administrative demands, supporting the organization of processes and analysis of cases with social impact. I am interested in applying technology, especially AI, to optimize workflows, expand access to information, and support vulnerable populations. I seek to develop solutions with real impact through data and innovation.