OPERATOR DOSSIER · 0001
Leandro Lugaresi
Backend & Systems Engineer
Backend systems that prefer to keep running when nobody is watching.
- Location
- Joinville · BR
- Status
- Open to ideas
- Role
- Senior backend
- Build
- OK
- Posts
- Pending
- Mode
- Dark / Clu
// 0001
About the
Operator
Backend engineer with 13+ years building high-performance data systems, APIs, and event-driven architectures. Core expertise in Go and TypeScript/Node.js, with deep experience in data pipeline optimization, distributed messaging, and platform infrastructure.
Currently owns a 6.4-billion-record TV/Film/Sports content catalog at BEN — daily incremental sync pipelines processing ~100K updates that used to take hours and now finish in under ten minutes. Open source contributor (148+ GitHub stars). Speaker at Gophercon Brazil 2016 and 2017.
— End transmission segment
// 0002
Telemetry Snapshot
3 readings · cached @ build
// 0003
Service Record
3 postings · 13+ years
- Posting 01
Branded Entertainment Network
Senior Software Engineer2021 — presentUtah, USA · RemoteOwns the core data integration layer for a TV / Film / Sports content catalog platform — 6.4B records (3.8B linear ratings + 2.6B streaming) ingesting from Luminate, Nielsen, and Gracenote. Team of 6 engineers.
- Rebuilt the Snowflake → OpenSearch sync pipeline (own initiative). Migrated from full S3 export to delta-based incremental sync over Snowflake Streams, processing ~100K updates daily. Sync time: hours → under 10 minutes.
- Built the Snowflake infrastructure from scratch — migration system, function/procedure sync, CLI for users + environments, RSA key pair setup, zero-copy cloning for dev/staging.
- Built a custom Snowflake MCP server (TypeScript) for AI-assisted debug queries and data exploration. Improved query accuracy across the whole codebase.
- Posting 02
LEVEE
Software Developer2019 — 2021São Paulo, Brazil · RemoteBuilt event-driven microservices for one of Latin America's largest HR platforms at the time (5–10M users), in Go, Ruby on Rails, RabbitMQ, and PostgreSQL.
- Rebuilt the RabbitMQ messaging topology across ~20 microservices. Three years of zero incidents after the redesign — directly led to open-sourcing the rabbids Go library.
- Replaced Google geocoding APIs in under a month when pricing changed, saving ~$15K/month (~$180K/year). Built the replacement using Brazilian public address data and Who's On First spatial data.
- Maintained Ruby microservices alongside implementing new Go services for the platform.
- Posting 03
Coderockr
Developer2014 — 2018Joinville, BrazilConsulting engineer embedded with client teams. Built greenfield systems for startups and improved reliability of existing platforms. Early Go adopter in Brazil — started learning Go here, which led to two Gophercon Brazil talks.
// 0004
Open Modules
3 repos · MIT
hub
High-performance Message/Event Hub for Go using publish/subscribe. Topics with RabbitMQ-style exchanges, no allocations on the hot path.
rabbids
Go wrapper around the AMQP client for RabbitMQ. Reconnect with exponential backoff, channel APIs, delayed-message support, and middlewares.
typesafe-config
TypeScript library for type-safe configuration. Pulls from multiple sources, merges them, validates against a schema. The boring config layer you actually want.
// 0005
Broadcast Log
2 talks · 2016 — 2017
Pipelines em Go
Concurrency patterns and pipelines in Go: how to use them efficiently across IO and CPUs, and how their use reduced a legacy data export from 5 hours to 7 minutes.