Bruno – React, Typescript, LLM

A biomedical data discovery and integration hub powered by AI agent automation. The platform unifies clinical and multi-omics datasets using FHIR, MEDS, and Croissant standards, accelerating sepsis research through improved data accessibility and democratization.

• Designed and implemented a three-layer platform architecture integrating Discovery, Harmonization, and Analysis capabilities, unifying heterogeneous clinical and multi-omics datasets through industry-standard formats. • Built a multi-interface system including a web UI for researchers, Python libraries for data scientists, and an MCP layer for AI agents. • Built automated data harmonization workflows preserving semantic context from GEO, ArrayExpress, and PhysioNet. • Developed containerized Nextflow analytical pipelines accessible via browser and programmatic interfaces. • Implemented a three-tier data access model (public, restricted, federated) incorporating differential privacy.

AI-powered embryo classification platform for cattle IVF, helping embryologists assess embryo quality using deep learning models.

• Built the complete platform from scratch, including admin dashboard, embryologist application, RESTful API, and real-time video processing pipeline. • Achieved a 68% reduction in video processing time (13.2s → 4.2s) through algorithm optimization and parallel processing. • Developed a multi-score frame selection algorithm using Laplacian variance, circularity scoring, texture analysis, and KMeans clustering. • Implemented a dual deployment strategy: Azure cloud for demos and on-premise GPU-accelerated servers for clinic production. • Designed a feedback collection system enabling embryologists to correct classifications for model retraining.

Gamified AI-assisted therapy activities platform for autism and other developmental conditions

• Scaled Fofuuu Edu from 500 to 2,500+ therapy activities through advanced internal tooling and standardized content production. • Secured R$1.5M+ in funding from Bossa Nova, Samsung, and Finep; led team expansion and product diversification. • Led a clinical study with Unimed demonstrating a 17% average increase in skill acquisition speed (31% peak). • Created the Fofuuu for Clinics B2B platform with AI-driven therapy plan customization and progress tracking. • Developed an automated activity metrics system capturing eight new metrics previously unobservable without digital assistance. • Implemented LGPD and HIPAA compliance and integrated OpenAI and LangChain for AI-powered performance insights.

Gamified innovative loyalty program project at Brazil's largest digital bank.

• Conducted comprehensive project analysis and realigned the MVP strategy with business objectives. • Led prototype creation and user validation testing, incorporating real user feedback before launch. • Coordinated with product designers and stakeholders to integrate the MVP into the existing product ecosystem. • Translated business requirements into technical user stories and development specifications.