ProjectsPixis — Ad-Tech Platform Engineering
Past — Pixisarchived
Pixis — Ad-Tech Platform Engineering
Multi-year backend engineering across Pixis's ad-tech platform — targeting-AI, performance models, reporting, sync pipelines, and upload infrastructure feeding Google, Meta, and Bing Ads.
Senior Backend Engineer
Jan 2019 – Dec 2023
What I built
- Targeting AI — audience inference service across social and search platforms
- Performance Models — ROAS / CPA prediction at campaign level
- Aiquire Reporting Backend — cross-platform reporting aggregator with sub-minute freshness
- Sync Engine + Upload Pipeline — Bing Ads, Google Ads, and Meta ingestion
- Execution Filter Engine + Token Management System — internal campaign orchestration
- Frequent-sync + performance audit services — keeping data fresh across providers
Hard problems
- Reconciling wildly different ad-platform schemas into one canonical reporting model
- Keeping the targeting-AI inference service sub-second at peak ad-server traffic
- Token lifecycle across 5+ ad platforms without a single-point-of-failure
Tech stack
PythonDjangoFastAPIPostgreSQLMongoDBRedisCeleryKubernetesAWSDocker
Tags
Ad-techTargeting AIPerformance MLReportingData pipelines
Work bundled across multiple internal services: targeting-ai, performance-models, aiquire-reporting-backend, sync-engine, upload-pipeline, execution-filter-engine, token-management-system, frequent-sync, performance-audit-service.
Source code is not in the public domain. Happy to walk through architecture or specific modules on a call — get in touch.