Work
Infrastructure work, data systems, and engineering challenges — with the metrics that actually matter.
Oracle Data Integrator constraints and a complex target API model caused the publishing flow to Idealista to degrade over time, creating large queues and operational issues for agencies.
Role: Designed the solution, built the MVP, led delivery (team of 3)
Impact
→97% performance improvement
→Queues reduced from ~8h peak to max 10m
→Per-listing time ~50s → ~1.5s
The legacy mailing system relied on PL/SQL procedures and scheduled jobs, topping out at ~12 emails/minute and creating severe delivery delays during backlog peaks.
Role: Proposed and implemented end-to-end
Impact
→Throughput ~12 emails/min → 1000+ emails/sec
→Eliminated peak-delay scenarios (days → near real-time)
→Preserved spooling pattern while scaling execution
Ingest cadastral data (Agenzia delle Entrate – Sister) delivered as files via S3, transform it through complex ETL, distribute JSON outputs to agencies, and integrate into Salesforce to automate property/client creation and updates.
Role: Owned the process end-to-end (first project with full decision ownership)
Impact
→Automated CRM registration of properties & clients
→Reduced manual entry time for agencies
→Greenfield pipeline built from scratch
The agency master-data delivery to external partners was fragmented and hard to maintain: per-partner tables/views/procedures/triggers plus layers of synonyms/DB links made lineage and reliability difficult.
Role: Designed the new centralized model and coordinated stakeholders
Impact
→Centralized architecture reduced complexity significantly
→Standardized source model with partner-specific outputs
→Improved maintainability and reduced redundant transfers
WSO2 slowed API development and depended on an externally managed system with limited in-house expertise. A Python-based approach was needed to regain speed and control for internal use cases.
Role: Proposed the migration, analyzed endpoints/traffic, helped design target architecture
Impact
→Reduced API delivery time by shifting to Python-first development
→Sized target infrastructure from traffic + volume analysis
→Improved internal control over API implementation
See the infrastructure I've built.
Interested in working together or learning more about a specific project?
Get in touch