Skip to main content
~/shantanu
Available for work

Shantanu Bhusari

I build backends that solve the right problem at the right scale. Go, distributed systems, event-driven microservices — 3+ years in production across platforms that had to handle real traffic. I care about what the product needs to do, and I make sure it can be shipped safely.

GitHub LinkedIn Resume
// 06Experience

Full-Stack Software Engineer

PayPilot Technologies LtdMar 2026 – Jun 2026 · Remote, India
  • –Turned plain-English queries into live financial reports — NL2SQL pipeline (Gemini 1.5 Pro + OpenAI) lets non-technical users query production data without SQL knowledge, across a multi-tenant Go/React 19 platform.
  • –Made email delivery a non-event: AES-encrypted BYOE SMTP + outbox pattern with MySQL SKIP LOCKED ensures zero message loss even when the downstream mail provider drops.
  • –Shipped to GCP with zero downtime — Terraform-managed Cloud Run + Cloud SQL, deployed via GitHub Actions; infra changes go out without taking the platform offline.
  • –Replaced manual report exports with PilotView BI Hub — users query, filter, and download CSV/XLSX/PDF in seconds; queries are validated against a schema allowlist before execution to prevent arbitrary SQL.

Lead Backend Engineer

Leifii MediaJan 2024 – Feb 2026 · Remote, India
  • –Designed comms strategy across 35+ services: gRPC for latency-sensitive internal calls, SQS for async workloads — deliberately chose SQS over Kafka to keep operations manageable at this scale.
  • –Traffic spikes on bulk order days used to spike compute costs — offloaded media transcoding and batch jobs to 15+ Lambda functions, keeping API response times stable without over-provisioning.
  • –5,000+ order and payment events processed daily with zero message drops — explicit SQS state transitions (payment_confirmed → shipped → invoice_queued) made async fan-out to 4 downstream services reliable.

Backend Developer

Northern AgencyApr 2023 – Jun 2024 · Remote, India
Node.js
  • –Shipped 3 production SaaS backends in 14 months across automotive, fitness, and logistics — chose monolithic Node.js/MongoDB deliberately; the scale didn't justify microservices overhead at that stage.
  • –Eliminated manual invoicing: automated PDF generation via Puppeteer + S3 replaced a fully manual process — clients went from hours of admin per week to zero.
  • –Logistics queries were hitting 60s+; rewrote MongoDB aggregations with compound indexes, dropped to sub-10s — the slow queries were forcing over-provisioned compute, so fixing them cut cloud costs 60%.

B.Tech — Electronics & Telecommunication

Vishwakarma Institute of TechnologyAug 2020 – May 2024 · Pune, India
  • –Focus on distributed systems and algorithms
  • –Built 10+ personal projects using JavaScript, TypeScript, Node.js, Express.js, and MongoDB
// 02Skills
Backend
Primary stack across all production systems
Go
PythonPython
Node.jsNode.js
gRPC
GraphQL
WebSocket
Frameworks
Web frameworks used across production services
FastAPI
Express.js
Hono.js
Databases
PostgreSQL + MySQL for relational · MongoDB for flexible schemas · Redis for caching
PostgreSQL
MySQL
MongoDB
Redis
RabbitMQ
Kafka
AWS
Serverless-first · Lambda + SQS across 3 platforms
AWS Lambda
AWS SQS
AWS SNS
AWS S3
AWS EC2
AWS ECS
AWS EventBridge
GCP
Cloud Run · Cloud SQL · zero-downtime CI/CD via Terraform
Google Cloud
DevOps
Containerisation, IaC, and CI/CD pipelines
Docker
Terraform
GitHub Actions
MinIO
Tools
Daily drivers for development
Claude Code
Cursor
Frontend
Next.js · supporting role in full-stack deliveries
Next.jsNext.js
React
TypeScript
Currently Learning
Experimenting — not yet in production
Rust
KotlinKotlin
Jetpack Compose
ScyllaDB
Kubernetes
← Back to home