Rudra Narayan Panda 0xrnp

Rudra Narayan Panda

Backend engineer who makes systems faster, cheaper, and harder to break.
_{Go · Rust · AWS · Distributed systems}
_{SWE L2 @ ASBL · Hyderabad}

LinkedIn · X · Dotfiles · LeetCode

RNP(1)                  Systems & Backend Manual

NAME
    0xrnp - latency hunter, terminal minimalist

DESCRIPTION
    Builds high-throughput Go microservices and distributed
    cloud infrastructure. Started in chemistry, pivoted to
    computer applications, now obsessed with storage engines,
    write paths, and cutting database bills without touching
    latency.

ENVIRONMENT
    Rex (Superlogical's terminal multiplexer) + Neovim

STACK
    Go, Rust, TypeScript, AWS (ECS, SQS, EventBridge),
    MongoDB, PostgreSQL, Elasticsearch

HISTORY
    ASBL (current) > Sambuq > Inncircles

Production impact

# WAL COMPACTION LOG (selected production overwrites)

  [asbl.mongodb]
- Atlas M60 burning compute on unindexed collection scans
+ Rewrote queries + compound indexes
+ Stepped down to M30: ~55% cheaper, p95 < 50ms

  [asbl.infra]
- Ad-hoc cron jobs drifting + manual financial reconciliation
+ Moved to ECS Scheduled Tasks
+ Automated EventBridge/SQS reconciliation pipelines

  [sambuq.search]
- Node.js hot path choking at 1.8s catalog search
+ Rebuilt in Go: 3x throughput
+ Elasticsearch n-grams: p95 < 80ms

  [sambuq.payments]
- Webhook race conditions, dropped state transitions
+ Idempotent Razorpay state machine + SQS DLQs

  [inncircles.ops]
- Multi-hour MTTR on production incidents
+ Standardized telemetry + runbooks: MTTR < 20 min

Public work

Most of my production work lives in private repos. What you can read and run today:

dotfiles: the Rex + Neovim setup I use every day.

Currently

# /proc/memtable (in memory, not fsynced yet)
[unflushed]
rust          = "the compiler is winning"
cassandra     = "reading the LSM-tree write path & compaction"
surrealdb     = "poking at storage primitives"
open_question = "what happens when a node dies mid-write?"

Ask me about

Query tuning and index design · Go service performance · idempotent payment and webhook flows · cutting database costs without hurting latency · incident response and runbooks

Say hi

Open to conversations about backend, storage, and infra problems. Best reached on LinkedIn or X.

_{"Sacrifice is the only way!"}

Navigation Menu