Rodovia 2026 · Pit stop
Luis Didonato
Software Engineer · Backend, AI & Cloud
CTO / Tech Lead at EVIGSão Paulo, Brazil
Software engineer with experience in backend, architecture and AI products. Background in technical leadership, integrations and cloud, focused on operational efficiency and scalability.
Impact
−80%
operating cost in the GCP → AWS migration
EVIG
2 s → 0.5 s
average load time on selected projects
RedBit
85.4%
mAP@50 on the model
Oceans-20
Tech Provider
WhatsApp Tech Provider approval
EVIG
Experience
Career path
EVIG
CTO / Tech Lead – Present
- Cut operating costs by 80%, even while taking on new projects, by migrating from GCP to AWS, combining Lambda with always-on services that scale independently, and optimizing containers.
- Planned the approval as a WhatsApp Tech Provider and led the review of apps, permissions, onboarding and WABA structuring.
- Designed an LLM-first architecture with agents, routing, conversational memory and decoupled services.
- Planned capacity and load tests to meet SLAs and prioritized the team’s deliverables.
- Planned and implemented RBAC and AWS SSM (configuration and secrets); planned JWKS.
Colmeia
Senior Integration Analyst –
- Planned integrations with partner companies and helped define flows and technical requirements.
- Implemented and validated integrations with partner and Google APIs, including Google Calendar.
- Planned improvements to Jira workflows to organize demands and track team deliverables.
ValiantGroup
Full Stack Developer –
- Built web and Android TV applications for the streaming platforms of the Hospital Sírio-Libanês, Housi and Fiesta Americana projects.
- Supported e-Saúde (Band) and the company’s AI chatbot, with Conversational Agents, a vector database and planning on Vertex AI.
RedBit
Front-end Developer –
- Built and optimized websites with a focus on responsiveness and SEO, cutting average load time from roughly 2 s to 0.5 s on selected projects.
- Set up tags, events and conversions with Google Tag Manager and Google Ads.
Cases
Selected work
LLM-first architecture
EVIG: agents, routing, conversational memory and decoupled services.
- LLMs
- Agents
- Routing
- Conversational memory
GCP → AWS migration
EVIG: −80% operating cost by combining Lambda with always-on services that scale independently.
- AWS
- GCP
- Lambda
- Docker
WhatsApp Tech Provider
EVIG: review of apps, permissions, onboarding and WABA structuring.
- WhatsApp Business
- WABA
- Onboarding
- Permissions
Oceans-20
Backend, frontend and mobile built end to end to route waste records to collection partners; model with 85.4% mAP@50.
- React Native
- FastAPI
- Firebase
Bicycle inspection (Porto Seguro)
Project built end to end, including an image-recognition microservice using YOLO.
- Java
- JPA
- React
- Oracle SQL
- YOLO
Workshop
Skills
Backend & Data
- Python
- Go
- Java
- C#
- FastAPI
- Django
- NestJS
- PostgreSQL
- SQL/NoSQL
Frontend & Mobile
- React
- Next.js
- TypeScript
- React Native
Cloud & Architecture
- AWS
- GCP
- Docker
- CI/CD
- Distributed systems
- Kafka
- RabbitMQ
AI
- LLMs
- RAG
- Embeddings
- Vector databases
- Multi-agent systems
Tools
- Claude Code
- OpenAI Codex
- Jira
Services
How I can help
LLM products
Agent architecture, RAG and memory, from prototype to production.
Cloud & cost
Migration, per-service independent scaling and load testing for SLAs.
Integrations
WhatsApp Business, partner APIs, Google Workspace.
Contact
Let’s talk
- gutodidonato [at] gmail [dot] com
- GitHub
- github.com/gutodidonato
- Résumé
- CV (PT) · PDF
- Languages
- English: advanced / C1 (EF SET)
- Spanish: intermediate