About · The architect
Architecture first. Always.
Origin
I'm Jorge Guberte — a software architect in São Paulo with twelve-plus years shipping production systems, now fully focused on applied AI systems architecture.
Most of the industry optimizes the moment: better answers inside a single session, then total amnesia. I work on the other axis — what happens to an AI system across weeks, months, years. How it accumulates understanding. How it forgets well. How it stays itself. And I hold that research to a production standard: it has to ship, serve real users, and survive real institutions.
The through-line
The work splits into a workshop and a lab. In the workshop: Strata, a B2G municipal intelligence platform running in production at Hub Esfera; Pixie, an embodied learning companion testing the research in the hardest arena there is — a child's attention. In the lab: LoomDB, memory engineered as a formally bounded dynamical system; EPCG, a categorical foundation for persistent agents whose memory access model already carries blind empirical validation; Ayvu-Talian, language modeling where the stakes of forgetting are cultural.
None of these are side projects. They are one program with multiple instruments — and every claim that matters gets a benchmark.
The practice
Before the AI chapter: nine years of foundations — distributed signage delivery, high-traffic campaign platforms for Walmart and Sony, cloud-native Python, Nike's global Joga3/JogaBonito campaign. That decade is why my AI work looks like systems architecture rather than prompt collections.
Today I'm co-founder and CTO of Hub Esfera, a GovTech company where I architect AI for complex public-sector data. Production is a forcing function: it keeps the research honest, and the research keeps production ahead of the curve.
The path so far
Sep 2024 — Present
São Paulo, Brazil · Hybrid
Co-Founder & Principal AI Architect · Hub Esfera
Architected Strata, a municipal intelligence platform for the public sector (B2G) with an embedded AI copilot — deployed on Cloud Run, serving production city profiles with live data pipelines and agent-driven decision support.
Jan 2023 — Present
Remote
Lead Applied AI Architect · Independent AI Systems Research
Ongoing applied R&D across agent memory, state management, generative UI, and multimodal AI — from experimental architectures to production-ready open-source libraries.
2023 · 5 months
Campinas, Brazil · Remote
Senior Developer Analyst (Contract) · Bosch Digital
Enterprise data integration and infrastructure modernization across complex SAP ecosystems, backend services, and distributed systems.
2006 — 2015
Brazil
Software Engineering Foundation · SeePix · Martin Luz · EgypTeam · Interactus
Nine years of production engineering before AI: distributed delivery platforms, high-traffic campaign sites, cloud-native Python, and interactive media for global brands.
Toolbox
- Languages
- Python · TypeScript · JavaScript · Rust (systems integration)
- AI systems
- Agentic workflows & orchestration · Typed memory interfaces & temporal retrieval · Advanced RAG · Context engineering · Model Context Protocol (MCP) · Agent observability, benchmarks & behavioral QA
- Frameworks
- Mastra · FastAPI · Node.js · Next.js · React · LangChain/LangGraph · CrewAI · AI SDK · BullMQ
- Data & storage
- PostgreSQL · Redis · Qdrant · Memgraph (Cypher) · DuckDB · Cloudflare R2
- Cloud & infra
- GCP Vertex AI · AWS Bedrock · Azure OpenAI · Modal serverless GPU · Docker · GitHub Actions
Where I operate
- 01Agent memory & temporal context architectures
- 02Typed retrieval interfaces beyond flat top-k
- 03Agentic orchestration for long-running applications
- 04Context engineering with verifiable evidence trails
- 05Generative UI & human-AI interaction
- 06Accessibility-first, neurodivergent-first design
If you're building something that needs to last longer than a session — or hiring someone who treats AI as architecture rather than incantation — we should talk.