Junior Software Engineer
Way2News · Jun 2023 – Present
- Built entire internal platform from scratch (2-person team) 80+ SPA screens, 440+ REST APIs, 8 user roles with a polyglot persistence layer across PostgreSQL, MySQL, MongoDB, ClickHouse, Redis, and SQLite, with RBAC and JWT-based auth throughout.
- Designed Kafka pipelines (KafkaJS) with multi-threaded producer/consumer services processing impression and click streams persisted into PostgreSQL, MongoDB, Redis, and ClickHouse; Redis caching layer for sub-millisecond content delivery.
- Built nightly aggregation pipelines (Node.js cron jobs) powering 30+ reports across operations, finance, and planning; offloaded heavy queries to ClickHouse; replaced manual spreadsheet tracking with real-time automation reducing operations team dependency by 40%.
- Built a Python Flask microservice integrating Google Ad Manager (SOAP API) and AdMob (OAuth2) report pipeline (DataDownloader CSV pandas), inventory configuration, order/line item creation, cloning, and bulk updates from the internal console.
- Implemented a Maker-Checker approval system enforcing four-eyes principle on all configuration and pricing changes before going live, with a full audit trail on every state transition.
- Delivered a Cashfree payment microservice server-side order creation, PostgreSQL persistence, HMAC webhook verification, server-to-server settlement callbacks, and background polling for real-time payment status reconciliation.
- Built Redis-backed listings API and embeddable widgets using Next.js (Edge Runtime) with parameterized URLs, form submissions, Cashfree.js checkout integration, geo-targeted priority sorting, and token-based attribution.
- Optimized the system with pre-materialized data for high-QPS reads handling ~150K requests/minute, achieving 95% inventory coverage and contributing to a 10% revenue increase.
- Designed and built a centralized BI platform with LLM-powered NL-to-SQL query generation, Debezium CDC + Kafka + ClickHouse materialized view pipeline, multi-tenant RBAC UI, and CSV/XLSX export consolidating analytics across all internal projects.

