BACKEND — DEVOPS — OPEN SOURCE
Raunak Kumar Jha
Backend Engineer | Distributed Systems | Open source contributor
Who I Am
I BUILD BACKENDS.
I BEND APIs TO MY WILL.
I AUTOMATE INFRASTRUCTURE.
I SHIP TO THE CLOUD.
I CONTRIBUTE TO OPEN SOURCE.
Career Timeline
2024
High School
Om Landmark School — PCM, 88%
2024
B.Tech Computer Science
Gandhinagar University, batch of 2028
2025
IIT Guwahati Micro-Credential
Micro-credit program in Computer Science & Engineering
2025
SDE Intern — Backend
Tech Mahindra — REST APIs, Redis caching, query optimization
2026
Software Dev Intern
Mangalam Information Technologies — backend & DevOps
2026
Open Source
Public contributions — backend & DevOps tooling
Now
Open to Opportunities
Backend-centric full stack, DevOps, exploring Web3
2024
High School
Om Landmark School — PCM, 88%
2024
B.Tech Computer Science
Gandhinagar University, batch of 2028
2025
IIT Guwahati Micro-Credential
Micro-credit program in Computer Science & Engineering
2025
SDE Intern — Backend
Tech Mahindra — REST APIs, Redis caching, query optimization
2026
Software Dev Intern
Mangalam Information Technologies — backend & DevOps
2026
Open Source
Public contributions — backend & DevOps tooling
Now
Open to Opportunities
Backend-centric full stack, DevOps, exploring Web3
Tech Stack
The stack that ships to production — each subsystem loads in sequence, like a clean boot.
Projects
01 — Case Study
Ledger
Real-time transaction ledger for a multi-tenant payments platform
Problem
Every write needed strict ordering and full auditability across tenants, but a single Postgres instance was buckling under write contention during peak load.
Solution
Partitioned writes by tenant through a Kafka-backed event log, replayed into per-tenant materialized views, with idempotency keys guaranteeing exactly-once application.
Challenges
- Guaranteeing ordering across partitions without a global lock
- Backfilling 18 months of history with zero downtime
- Keeping read replicas under 200ms of replication lag
Stack
Pipeline
14.2M
Events / day
38ms
P99 latency
99.97%
Uptime
02 — Case Study
Meshline
Service-mesh observability layer for a 60-microservice fleet
Problem
Incident response took 40+ minutes because no single view connected request traces, deploys, and infrastructure metrics.
Solution
Built a correlation layer that stitches OpenTelemetry traces to Grafana dashboards and deploy events, surfacing the likely root cause automatically.
Challenges
- Sampling traces at scale without losing the rare failure paths
- Correlating deploy timestamps across independently-shipped services
- Keeping the ingestion pipeline cheaper than the incidents it prevents
Stack
Pipeline
-62%
MTTR reduction
60
Services covered
9.1K
Traces / sec
01 — Case Study
Ledger
Real-time transaction ledger for a multi-tenant payments platform
Problem
Every write needed strict ordering and full auditability across tenants, but a single Postgres instance was buckling under write contention during peak load.
Solution
Partitioned writes by tenant through a Kafka-backed event log, replayed into per-tenant materialized views, with idempotency keys guaranteeing exactly-once application.
Challenges
- Guaranteeing ordering across partitions without a global lock
- Backfilling 18 months of history with zero downtime
- Keeping read replicas under 200ms of replication lag
Stack
Pipeline
14.2M
Events / day
38ms
P99 latency
99.97%
Uptime
02 — Case Study
Meshline
Service-mesh observability layer for a 60-microservice fleet
Problem
Incident response took 40+ minutes because no single view connected request traces, deploys, and infrastructure metrics.
Solution
Built a correlation layer that stitches OpenTelemetry traces to Grafana dashboards and deploy events, surfacing the likely root cause automatically.
Challenges
- Sampling traces at scale without losing the rare failure paths
- Correlating deploy timestamps across independently-shipped services
- Keeping the ingestion pipeline cheaper than the incidents it prevents
Stack
Pipeline
-62%
MTTR reduction
60
Services covered
9.1K
Traces / sec
Open Source
- init: project scaffold
- feat: add JWT auth middleware
- feat: redis caching layer
- test: cache invalidation edge cases
- fix: race condition in cache writer
- merge: redis-cache into main
- chore: bump dependencies
- perf: connection pooling
raunak0400
PythonDynamic Realtime profile ReadMe linked with Spotify.
Flask-CXR
PythonCXR (Sixer) - Comment-Driven Code Editor A web-based code editor that converts comments into functional code, analyzes existing code for errors, and supports multiple programming languages. 🚀
Hospital-Medical-Information-System
JavaScriptNo description provided.
similarity-checking-from-books
C++No description provided.
Contribution Activity
Interactive Terminal
Try: help · projects · skills · theme