Scaling Data Pipelines & Growth Analytics
•
AirflowAWSReact-NativePythonFastAPIML Sentiment Models
Overview
At A&C Future, I worked on large-scale ETL pipelines and analytics systems serving 6M+ user records, while also shipping user-facing React-Native features with integrated event tracking.
The Challenge
- Data integrity issues across millions of records
- User behavior analysis requiring detailed event tracking
- Community perception analysis from 30k+ social media comments
Architecture Decisions
Airflow-Powered ETL
- Designed DAGs to ingest and transform data daily.
- Improved coverage and reliability, eliminating prior integrity issues.
React-Native Analytics Integration
- Built components with event tracking hooks.
- Enabled growth, retention, and behavior analysis at scale.
ML-Based Sentiment Analysis
- Applied transformer models to classify 30k+ comments from Telegram and Twitter.
- Extracted actionable insights on engagement and brand perception.
Key Learnings
- Balancing ETL reliability with analytics accuracy is key in fast-growth environments.
- Event instrumentation directly empowers growth decisions.
Metrics
- 6M+ user records processed
- 30k+ social media comments analyzed
- Analytics dashboards adopted company-wide