Open to work

Badri Vishal Padhy

AI Backend Engineer · Noida, India

I build backend systems for AI agents and distributed products — event-driven pipelines, reliable test infrastructure, and APIs that hold up under failure.

I mostly work in Python and TypeScript, building services with FastAPI and Node.js, backed by PostgreSQL and Redis, wired together with Kafka, and shipped with Docker on AWS.

Experience

  1. AI Backend Engineer

    TestMu AI
    PythonFastAPIPostgreSQLKafkaAI AgentsLow-Level Systems
    1. Developed backend features for agent-to-agent testing across chat, voice, and phone agents.
    2. Fixed evaluation settings and improved AI prompts to make test scenarios clearer and test results more reliable.
    3. Improved test-run tracking and handled duplicate updates and timeouts across chat, voice, and phone testing.
    4. Built Slack and email alerts for test outcomes; enforced organization-scoped access to protect agent context, evaluation data, and LLM billing.

  2. Software Developer

    Litekart
    Remote
    React.jsSvelteHono.js
    1. Built frontend features and product modules using React.js, Svelte, and Hono.js in a 6-member team, delivering features from frontend to backend.
    2. Improved page load time by 25% by fixing performance issues in the rendering pipeline.
    3. Integrated REST APIs to reduce client-side errors and cleaned up code across multiple services.

Projects

  1. Nodex · Distributed workflow automation platform

    LiveGitHub
    Next.jsExpress.jsKafkaPostgreSQLDocker
    1. Architected a Zapier-like platform with 5 decoupled microservices using Kafka for event-driven workflow orchestration.
    2. Implemented a PostgreSQL transactional outbox and Node.js publisher worker for at-least-once delivery under failure scenarios.
    3. Dockerized services with multi-stage builds and deployed with CI/CD for automated testing and container releases.

  2. Exness · Crypto futures trading platform

    GitHub
    TypeScriptNode.jsRedis StreamsWebSocketsPrismaPostgreSQL
    1. Built an asynchronous trading engine with Redis Streams consumer groups, ACK-based confirmation, and retry logic.
    2. Ingested and broadcast live market data at approximately 10 updates/sec through WebSockets.
    3. Implemented order placement, matching, and settlement with PostgreSQL persistence in a typed Turborepo monorepo.

  3. Twixy · Twitter-like social platform

    GitHub
    Node.jsGraphQLPostgreSQLRedisNext.jsTypeScriptAWS
    1. Built GraphQL APIs for tweets, follows, likes, and personalized feeds with cursor-based pagination.
    2. Cached sessions and feeds in Redis; used Next.js SSR/ISR for fast initial page loads and public-page SEO.
    3. Deployed on AWS EC2 with S3 media storage and managed PostgreSQL migrations using Prisma ORM.

Skills

Languages
TypeScriptJavaScriptPythonJavaGo
Frontend
Next.jsReact.jsSvelteTailwind CSS
Backend
Node.jsExpress.jsFastAPIPostgreSQLRedisPrisma ORM
Systems & DevOps
KafkaRedis StreamsWebSocketsMicroservicesDockerAWSTurborepoCI/CDLinuxGit

Education

XIM University · B.Tech in Computer Science

Achievements

Hacktoberfest 2024 Contributor

Contributed multiple successfully merged pull requests to open-source repositories.

Activity

Loading GitHub activity…