Contributed to the CRM platform agency owners use to manage creators, with a focus on day-to-day communication workflows and operational tooling.

Designed and shipped a shared-worker chat widget for real-time inbox updates, typing states, connection status, and message actions across chat surfaces.

Worked across a TypeScript monorepo: Next.js/React frontends, Node/Bun services and workers, shared UI packages, PostgreSQL with Drizzle ORM, and event-driven job workflows.

Worked at Teamlease Digital on contract to Infosys Consulting, alongside engineers from Temasek, maintaining and extending their internal product suite.

Supported multiple Node.js API microservices and React frontends on Kubernetes, integrating shared JavaScript libraries from a private JFrog instance.

Used PostgreSQL as the primary database and Redis clusters for caching to improve responsiveness and reliability.

Maintained and extended the existing product while improving application UX. Core contributor to the second product iteration, working with NestJS, Redis, JupyterLab, and Kubernetes.

Helped move from a single CloudFormation stack to a distributed, self-hosted microservices architecture. Contributed to infrastructure with Istio, EKS, and ArgoCD, and implemented inter-service communication over Redis channels.

Built an approval-backed schema and data migration pipeline for Postgres (TypeScript migration scripts, GitHub Actions) that improved the developer workflow.

Part of the core team behind MEAT (Most Economically Advantageous Tender), a framework that helps users assess bid win likelihood.

Turned high-fidelity Figma designs into React UI with state patterns defined in design. Operated Elasticsearch, MongoDB, and DynamoDB datasets shaped for buyer and supplier intelligence at scale.

Increased reuse across several microservices using OOP, OOAD, and TDD practices. Stack included TypeScript, React, Node.js, Serverless, AWS Lambda, DynamoDB, MongoDB, and CloudFormation.

Designed and built an in-house interview and video calling platform to streamline hiring. The voice/video server used Jitsi; the code execution engine used Judge0.

Gained exposure to deployment and scaling with Nginx, CloudFormation, Route53, CloudFront, and PM2. Stack included TypeScript, React, Node.js, DynamoDB, and MongoDB.