$ whoami

Saurabh
Shukla.

AI Developer & Software Engineer . Fuelled by coffee
Learning and developing high-scale, production-grade systems and intelligent agents.

view work try playground →

Bengaluru, India open to relocate↳ B.E. CSE · 9.1 GPA

01.

about

I design production-grade systems that are observable, reliable, and ship fast.

Currently, I build and scale distributed systems that move billions of customer events daily. My work focuses on high-volume event ingestion, deep observability, and developing intelligent AI tooling—like MCP-based agents that translate natural language into complex SQL and data workflows.

My technical foundation is built on architecting distributed streaming pipelines and event-driven architectures (Kafka, Pub/Sub, RabbitMQ).

I’m a competitive programmer at heart—having solved 650+ DSA problems with a Global Rank of 200 on CodeChef. Whether it's winning Web-a-thon 2023 or maintaining 750+ GitHub contributions since 2025, I am obsessed with shipping reliable code, fast.

Web-a-thon 2023

Winner

HackDay-21

2nd Runner-Up

B.E. Computer Science

Chitkara University, Himachal Pradesh

Jul 2021 — Aug 2025 · GPA 9.1 / 10

/now

updated jul 2025

Current Focus
Distributed systems & AI-driven observability
Building
AI applications (IMO, DocuQuery) & high-scale data systems
OSS Contribution
750+ contributions since 2025 (GitHub)
Learning
HLD, LLD, & Backend Architectures
Reading
System Design Interview — Alex Xu

02.

experience

Zeotap

·Software Engineer(Prev: SDE Intern)

May 2025 — Present Bengaluru, IN

→Implemented scalable request-response logging and observability for distributed streaming and batch pipelines using Pub/Sub and BigQuery.
→Delivered 20+ partner integrations (Amazon DSP/AMC, StackAdapt, Zoho, Talon.One, YieldLabs) across ingestion and real-time pipelines for Zeotap's CDP.
→Developed the Policy Lifecycle Engine enabling rule-based data routing — clients configure rules and prioritize delivery across downstream systems.
→Built Integr8, an AI integration agent that scans developer documentation and auto-generates database configurations and test scripts — reducing setup time by 90%.
→Built Doctor-Integr8 (internal AI tool / MCP + Slack bot) that converts natural language queries into complex SQL and returns insights on ingestion lifecycle and historical estimations.
→Created automated Slack alerting pipelines detecting ingestion failures and RabbitMQ lag using Python and SQL.

JavaMavenDistributed SystemsRabbitMQGCPBigQueryGCSPub/SubPythonLangGraphLangChainFastAPIPostgreSQLAngularJSTypeScript

Summit Technology Group India

·Software Engineer Intern

Jun 2024 — May 2025 Chandigarh, IN

→Designed and developed the loan servicing and payment backend system from the ground up, with bank account onboarding/disconnection flows, recurring billing, and scheduled payment processing.
→Developed a cache layer using Django, increasing tenant database performance by up to 40% and reducing database load during peak hours which improved database scalability.
→Enhanced backend services by integrating third-party government APIs and orchestrating asynchronous workflows with Celery task queues and workers.
→Optimized CI/CD pipelines (GitHub Actions, CircleCI) by trimming log output and enabling parallel pytest execution, resulting in 50–60% faster pipelines.

DjangoPytestSQLDockerGitHub ActionsCircleCIJavascriptVue.jsjq

03.

selected projects

IMO — AI Shopping Assistant

Jan 2026

Distributed AI-powered product research platform aggregating products and reviews from multiple marketplaces in real time — collapsing hours of manual comparison into minutes.

Async task orchestration with Celery workers, task polling, and Redis-based progressive caching to keep long-running AI pipelines off the request thread.

Marketplaces aggregated

~85%

Latency reduction

PythonFastAPIPostgreSQLAWSReact.jsDockerRedis

DocuQuery

Oct 2025

Built an AI-powered chat-to-PDF system that enables users to edit PDFs via natural language in seconds while preserving original layout, styling, and formatting, unlike typical LLM tools.

Implemented vector-based retrieval and document-aware editing using embeddings for precise section-level updates without corrupting the document structure.

< 8s

Avg edit time

100%

Layout preserved

FastAPILangChainReactChromaDBDocker

04.

toolkit

Languages

PythonJavaJavaScript / TypeScriptSQLC++Dart

Frontend

React.jsVue.jsNext.jsTailwind CSSReduxSocket.io

Backend

FastAPIDjangoNode.jsExpressPlayCeleryPrismaHikariCP

Data & Distributed

PostgreSQLMySQLMongoDBRedisKafkaRabbitMQPub/Sub

Cloud & DevOps

AWSGCPDockerKubernetesGitHub ActionsArgo CDMaven

AI / Automation

LangChainLangGraphRAGMCPLLM WorkflowsAI Agents

Testing

PytestJUnitMockito

Python •Java •JavaScript / TypeScript •SQL •C++ •Dart •React.js •Vue.js •Next.js •Tailwind CSS •Redux •Socket.io •FastAPI •Django •Node.js •Express •Play •Celery •Prisma •HikariCP •PostgreSQL •MySQL •MongoDB •Redis •Kafka •RabbitMQ •Pub/Sub •AWS •GCP •Docker •Kubernetes •GitHub Actions •Argo CD •Maven •LangChain •LangGraph •RAG •MCP •LLM Workflows •AI Agents •Pytest •JUnit •Mockito •Python •Java •JavaScript / TypeScript •SQL •C++ •Dart •React.js •Vue.js •Next.js •Tailwind CSS •Redux •Socket.io •FastAPI •Django •Node.js •Express •Play •Celery •Prisma •HikariCP •PostgreSQL •MySQL •MongoDB •Redis •Kafka •RabbitMQ •Pub/Sub •AWS •GCP •Docker •Kubernetes •GitHub Actions •Argo CD •Maven •LangChain •LangGraph •RAG •MCP •LLM Workflows •AI Agents •Pytest •JUnit •Mockito •

05.

playground

A live terminal that knows me well. Try help, projects, or joke.

saurabh@portfolio:~

zsh

saurabh.dev playground v1.0.0

type 'help' to see available commands.

▍

06.

get in touch

Got an idea, role, or
just want to talk shop?

I read every message. Quickest way is email or LinkedIn — I usually respond within 24 hours.

+91 8149190804 /in/sv176734 @Saurabh13042004

SaurabhShukla.