Backend engineer by trade, systems thinker by habit. I've designed APIs handling a thousand requests a second, shipped a Python SDK with thousands of downloads and built AI agents that review code like a senior engineer would.
MoEngage needed a unified way to manage messaging templates across 6 channels — Push, Email, SMS, In-App, On-Site, and WhatsApp. No central system existed. I built one from scratch.
The system handles folder organisation, REST APIs, rate-limiting, Redis caching, and DB optimisations. It sits in the critical path of all outbound & inbound messaging.
Marketers were writing raw Jinja syntax to personalise campaigns. Error rates were high, campaigns were breaking, and non-technical teams were blocked entirely.
I built a visual content builder that abstracts all template logic into a drag-and-drop UI — generating correct Jinja under the hood automatically.
Enterprise clients were sending campaigns to their entire audience — millions of users — when the goal was a controlled rollout. No cap mechanism existed.
I designed dynamic audience capping enforced at the campaign level, before any message is dispatched.
Developers integrating with MoEngage Inform's transactional messaging (3B+ messages daily) had to deal with raw HTTP and auth flows. I designed and shipped a clean Python SDK to PyPI with proper error handling and versioned releases.
Also unified AWS and Azure deployments into a single CI/CD pipeline, cutting release overhead in half.
An AI-powered GitHub App that auto-triggers on pull request events and delivers codebase-aware code reviews. Uses a LangChain + Qdrant RAG pipeline to understand your full repo context — not just the diff — and supports back-and-forth feedback through an inline PR comment chatbot.
I'm a software engineer with 3 years at MoEngage — building systems that handle billions of messages and thousands of enterprise clients.
I care about clean APIs, observable systems, and shipping things that actually work at scale. Right now I'm most drawn to the intersection of backend infrastructure and LLM agents.
Open to interesting conversations — roles, collaborations, or just a good problem to talk through.