Hey, I'm

Pulkit Agrawal

Staff Software Engineer (IC4) @ Snowflake

Writing about scalable systems & AI-ready engineering. Over 9+ years, I've worked across fintech, crypto, mobility and data at companies like Snowflake, BitGo, PhonePe, Rapido and Tricon, designing systems that handle millions of users.

0+
Years Experience
0M+
Users Served (PhonePe)
0%+
Outage Reduction
0%
Cost Reduction
0M+
Daily Rides (Rapido)

Key Impact

Orchestrated distributed systems for core infrastructure supporting 400M+ users, focusing on authentication and notification services
Designed Project Change Management Request framework, reducing production outages by 50%+
Designed and developed Coupon Generation, Notification, and Subscription microservices handling 1M+ daily rides
Spearheaded migration from traditional VM infrastructure to Docker and Kubernetes, reducing costs by 40%

Featured Writing

View All →
Designing Agent-Ready APIs in the Real World
Medium

Designing Agent-Ready APIs in the Real World

Lessons learned from wiring legacy endpoints to LLMs

Read Article
Building an AI-Powered Interview Preparation Assistant (Part 2)
Medium

Building an AI-Powered Interview Preparation Assistant (Part 2)

Expanding Prep AI: Multi-Agent System for Holistic Interview Preparation

Read Article
Building a LeetCode Chatbot Using AI: A Comprehensive Guide
Medium

Building a LeetCode Chatbot Using AI: A Comprehensive Guide

DSA AI: Your Personalized DSA Companion!

Read Article
How I Built an AI Agent Over an Existing Backend (Using MCP, Without a Rewrite)
LinkedIn

How I Built an AI Agent Over an Existing Backend (Using MCP, Without a Rewrite)

They need a new interface layer that can: understand what a user wants, call the right backend capabilities safely, and not turn production into a prompt lottery. A working, end-to-end demo with Agent Service, MCP Server, and existing backend.

Read Article
Why Our "Normal" APIs Weren't Agent-Ready (and What We Changed as Backend Engineers)
LinkedIn

Why Our "Normal" APIs Weren't Agent-Ready (and What We Changed as Backend Engineers)

On a recent project, we tried something that sounded simple: Take our existing, battle-tested APIs and plug them into an LLM-based agent to automate workflows. These APIs were not side projects — they were used in production by multiple teams. Here's what we changed.

Read Article
Mastering Rate Limiting with the Token Bucket Algorithm: A Comprehensive Guide
Medium

Mastering Rate Limiting with the Token Bucket Algorithm: A Comprehensive Guide

Creating a Multithreaded Token Bucket Rate Limiting Algorithm

Read Article

Featured Projects

View All →

Why Subscribe Here?

Curated Case Studies

Real-world engineering stories with metrics and learnings

All Writing in One Place

Medium + LinkedIn articles centralized

Direct to Your Inbox

Subscribe once, never miss an article

Experience

Full Timeline →

Staff Software Engineer (IC4)

Snowflake Bengaluru, Karnataka, India
Apr 2025 - Present
  • Building AI-ready data infrastructure and backend systems at scale
  • Working on distributed systems for enterprise AI and ML workloads
  • Driving technical excellence and system design across teams
Artificial Intelligence Backend Distributed Systems Cloud Infrastructure

Staff Software Engineer (E3)

BitGo Bengaluru, Karnataka, India
May 2024 - Apr 2025
  • Led Global Travel Rule project ensuring seamless digital asset withdrawals across countries while complying with local regulations
  • Developed risk scoring service to enhance detection of money laundering and terrorist financing using Snowflake and Postgres
  • Contributed to KYC verification flow with Persona, integrating real-time data pipelines for status synchronization
TypeScript Apache Airflow PostgreSQL Snowflake Distributed Systems

Lead Software Engineer

PhonePe Bengaluru, Karnataka, India
May 2022 - May 2024
  • Led high-performing platform team driving development of scalable in-house platform products emulating cloud-like functionality
  • Orchestrated distributed systems for core infrastructure supporting 400M+ users, focusing on authentication and notification services
  • Architected PPEC Cloud API Architecture with APIs, RBAC, CLIs, and SDKs for provisioning cloud resources across data centers
  • Designed Project Change Management Request framework, reducing production outages by 50%+
  • Pioneered asynchronous architecture using Celery, RabbitMQ, and Airflow for improved scalability
Distributed Systems FastAPI Go Python Apache Airflow HBase Microservices Kafka

Publications & Recognition

See All →

VayuX AI

V

VayuX AI

Building AI-powered tools for the future

VayuX AI is focused on creating intelligent tools using AI, RAG systems, and modern agent architectures. Our flagship project Prep AI helps candidates prepare for technical and behavioral interviews.

Let's Connect!

Join my newsletter for insights on scalable systems, AI-ready engineering, and building impactful products.