•Designed, advocated for, engineered and shipped several consumer-facing backend API services to user applications (web3 clients) with a combined 30million+ MAUs including:
a) A multi-database query service (SQL, NestJS, Typescript, Elasticache)
b) An API service to fetch multi-account & multi-network balances and transaction history for users.
c) A token icons caching service (S3, K8s)
d) A phishing detection service (S3, K8s)
•Worked with the team to introduce improvements to high-volume API service latency and decrease reliance on 3rd party API services for data access.
•Introduced an extensible Prometheus Metrics service implementation & Grafana dashboards for ease of observability of individual subcomponents across 10+ backend services. Extended this to other internal teams to increase across-the-board observability breadth. Increased observability led to:
a) A ~20-30% overall cost improvement for our shared team's costs by customized tweaks to deployment settings across various service types.
b) Up to 90% improvements across various user-facing feature latency by exposing specific subcomponent (internal/external API endpoints, cache reads/writes, db reads, etc) latency slowness.
•Refactored multiple high-volume API deployment stacks to improve reliability & optimize costs using internal routing, request/response caches (AWS Elasticache) and improved edge caching layers (Cloudflare, primarily).
•Led a team of 4-5 engineers to accomplish quarterly objectives and worked with director of engineering and product manager to define the quarterly objectives.
•Led a documentation effort to improve internal docs site coverage of internally & externally-available API services.
