Work
Infrastructure work, data systems, and engineering challenges — with the metrics that actually matter.
Legacy batch pipeline couldn't handle peak traffic spikes, causing 4-6 hour data delays in downstream ML models.
Role: Lead Data Engineer
Impact
→Handled 12M+ events/day
→Reduced end-to-end latency from 4h to 90s
→99.97% uptime over 6 months
Data science teams were duplicating feature engineering logic across 8+ models, causing inconsistencies and wasted compute.
Role: Backend & ML Infrastructure Engineer
Impact
→Reduced feature computation time by 63%
→Unified 40+ features across 8 models
→Cut model deployment time by half
Migrating a 10TB legacy Redshift warehouse to Snowflake with zero downtime and full historical parity.
Role: Data Platform Engineer
Impact
→Migrated 10TB+ data with zero data loss
→Reduced query costs by 41%
→Zero downtime migration
Fragmented cron-based ETL scripts with no observability, retries, or dependency management — failed silently.
Role: Data Engineer
Impact
→Orchestrated 120+ DAGs in production
→Reduced pipeline failures by 78%
→Full lineage and alerting coverage
Unstructured documents (PDFs, emails, reports) contained key business data that was inaccessible for analysis.
Role: AI & Data Engineer
Impact
→Extracted structured data from 50K+ documents
→91% accuracy on key entity extraction
→Saved ~300 manual hours/month
See the infrastructure I've built.
Interested in working together or learning more about a specific project?
Get in touch