shekhar.os — Senior Software Engineer

SHEKHAR.OS v4.21_

--:--:--UPTIME 00:00

/usr/bin/whoami

shekhar@kumar:~$ ./introduce --mode=senior-engineer

 ____  _   _  _____  _  __ _   _    _    ____
 / ___|| | | || ____|| |/ /| | | |  / \  |  _ \
 \___ \| |_| ||  _|  | ' / | |_| | / _ \ | |_) |
  ___) |  _  || |___ | . \ |  _  |/ ___ \|  _ <
 |____/|_| |_||_____||_|\_\|_| |_/_/   \_\_| \_\

I build backends that don’t flinch under load.

Senior Software Engineer, 4+ yrs. I design data‑heavy platforms in Node · NestJS · Postgres, wire up AI/RAG pipelines, and obsess over clean service boundaries. Currently shipping Ivorie— a wedding marketplace w/ 5k+ users, 10k+ msgs/day.

LOC Bengaluru, INSTATUS ◉ open to opportunitiesTIMEZONE IST (UTC+5:30)

DOWNLOAD CV INITIATE CONTACT GITHUB

shekhar@kumar:~$

NAMESHEKHAR KUMARIDEMP_00235838CLEARANCELVL 4 / SR. ENGINEERLAST SEENjust now

Node.js · NestJS · TypeScript · PostgreSQL · Prisma · React · Next.js · React Native · LangChain · RAG · Vector DBs · AWS · Redis · StripeNode.js · NestJS · TypeScript · PostgreSQL · Prisma · React · Next.js · React Native · LangChain · RAG · Vector DBs · AWS · Redis · Stripe

about.md [01/07]

cat about.md

# hey.

I’m Shekhar — a senior backend engineer who’s spent the last four years building the unglamorous layer that makes products feel magical: schemas that don’t rot, queues that don’t drop, auth that doesn’t panic at 3am.

Lately I’ve been in the weeds on agentic AI— RAG pipelines, LangChain, fine-tuned chatbots, LLM-assisted testing. I care about shipping things that are measurable, A/B-tested, and boring to operate.

Off keyboard: I think quietly, ask more questions than I answer, and trust honesty over politeness.

nameShekhar KumarroleSenior Software Engineerexp4+ yrs · backend / full-stackfocusScalable systems, RAG, A/B testinglangTypeScript, Python (reading)stackNode · NestJS · PG · Prisma · NextaiLangChain · Vector DBs · LoRA tinkeringcloudAWS (SQS, S3, Lambda)shippedIvorie · FSAI · GiftCash · Weather.comcurrentlyDiffusion models + LoRA fine-tuning

about.md · 1.2 KB · read-onlyESC to exit · J/K to scroll

experience.log [02/07]

tail -f experience.log

2024 — PRESENT

CURRENT

Senior Software Engineer

Ivorie — Wedding Marketplace ● Full-stack · NestJS / React Native / Next.js

Owned backend architecture for a marketplace serving 5,000+ users; defined service boundaries, DB schemas, and event-driven async workflows.
Built real-time chat + notifications handling 10k+ msgs/day, plus Stripe payment flows and AWS-based media uploads.
Delivered Expo React Native app (iOS + Android) and a Next.js admin dashboard from zero to production.
Introduced analytics dashboardstracking 20+ product metrics — shifted roadmap from vibes to data.

2023 — 2024

Software Engineer — AI & Platform

FSAI / Weather.com / GiftCash ● React · Next.js · LangChain

Architected RAG pipelineswith vector DBs — improved response relevance by 25%.
Built a generative AI chatbot using LangChain with categorized prompts, fine-tuned models, and real-time web search via custom MCP server.
Led Fleet, an internal AI-driven platform inspired by GitHub + Booking.com — validated resource modeling before production rollout.
Ran A/B experiments on Weather.com (video duration, layout variants); improved engagement 15–20%.
Shipped product catalog module for GiftCash.com — cut manual admin ops 30%.

2022 — 2023

Software Engineer

Stocks Investment Platform ● Express · MongoDB · Redis

Built a platform letting experts publish strategies for retail investors — leaderboards for 1k+ users.
CronJobs for live stock price updates and return calculations.
AWS SQS + Twilio WhatsApp for real-time trade and basket notifications.

2021 — 2022

Software Engineer

Early career ● Node · React · MongoDB

Shipped production features across web & mobile; learned what a healthy deploy feels like.
Introduced LLM-assisted testing— user-story‑driven test generation + randomized edge cases, +35% coverage.
Built automated multi-tenant migration system — safe schema & feature propagation with zero downtime.

4 entries · 4 years · 1 timezone⇧ G to end of log

projects/ [03/07]

01 · PRODMARKETPLACE

┌─── ivorie ─────────────────────────┐ │ users ████████████░░░░ 5.1k │ │ msgs/day ███████████████░ 10.2k │ │ uptime ████████████████ 99.9% │ │ services █████░░░░░░░░░░░ 12 │ └────────────────────────────────────┘

Ivorie

Full-stack wedding marketplace. NestJS API, Expo app, Next.js admin. Real-time chat, Stripe flows, event-driven backend.

5k+users

10k+msgs/day

99.9%uptime

NestJSPostgresPrismaExpoNext.jsStripeAWS S3

/projects/ivoriePROD ●

02 · PRODAI / RAG

┌─── rag-chatbot ────────────────────┐ │ query → [embed] → [vectordb] → ctx │ │ │ │ │ ctx + history → [llm] → response │ │ │ │ relevance +25% · p50 latency 420ms │ └────────────────────────────────────────┘

RAG Chatbot + MCP

Generative AI chatbot with LangChain, categorized prompts, fine-tuned models, real-time web search via custom MCP server.

+25%relevance

4fine-tuned models

MCPcustom server

LangChainVector DBMCPPythonTypeScript

/projects/rag-chatbotPROD ●

03 · INTERNALPLATFORM

┌─── fleet ──────────────────────────┐ │ github-like navigation ✓ │ │ booking.com-style search ✓ │ │ ai-assisted workflows ✓ │ │ scalable resource modeling ✓ │ └────────────────────────────────────────┘

Fleet

Internal AI-driven platform, inspired by GitHub + Booking.com. Validates scalable resource modeling, search, and AI-assisted workflows before production rollout.

LEADproject owner

AI-nativecore loop

Next.jsNestJSPostgresLangChain

/projects/fleetINTERNAL ●

04 · PRODFINTECH

Stocks Investment Platform

Platform where experts publish strategies for retail investors. Live price updates via CronJobs, WhatsApp alerts via Twilio + AWS SQS.

1k+users

RealtimeSQS + Twilio

ExpressMongoDBRedisAWS SQSTwilio

/projects/stockxPROD ●

skills --tree [04/07]

proficiency report · self-assessed

backend

Node.jsNestJSExpressTypeScriptRESTGraphQLEvent-drivenMicroservices

data

PostgreSQLPrismaMongoDBRedisVector DBsSQSSchema designMigrations

frontend

ReactNext.jsReact NativeExpoTailwind

ai / ml

LangChainRAGMCPFine-tuningLoRADiffusion (learning)

cloud / ops

AWSS3SQSLambdaDockerCI/CD

practice

A/B testingAnalyticsLLM-assisted testsSystem designStripeTwilio

Node / NestJS

95%

TypeScript

93%

Postgres / Prisma

90%

React / Next.js

85%

React Native / Expo

80%

LangChain / RAG

82%

AWS (SQS, S3, λ)

78%

Diffusion / LoRA

45%

meters reflect comfort, not egolast calibrated: today

now --currently [05/07]

systemd status · this week

[mon 09:12] [OK] shipped multi-tenant migration system to prod · zero downtime ✓ [mon 14:40] [OK] instrumented new A/B on Ivorie discovery feed [tue 11:05] [RND] spent the afternoon reading the LoRA paper (again) [wed 10:22] [OK] prototyped diffusion fine-tune on local GPU · it converged, barely [wed 22:11] [HMM] refactored RAG retriever · scored relevance +4pp [thu 09:00] [OK] wrote LLM-generated edge-case tests for Ivorie booking flow [fri 16:30] [...] planning the next 90 days · less code, more systems [now] [●] open to senior backend / AI engineering roles

pid 1 · shekhar.osrefreshed live

years shipping

senior backend + AI

users reached

across Ivorie + platforms

msgs/day handled

real-time chat infra

coffee

critical dependency

contact.sh [06/07]

interactive shell · try typing: help

Welcome to shekhar.os · type help for commands. Available: help · about · skills · projects · contact · social · resume · coffee · sudo · clear

➜shekhar.os$

EMAILshekharkumar235838
@gmail.com WHATSAPP+91 84092 09072 GITHUB@HITMANdark07 LINKEDIN/in/shekhar
-kumar1

encrypted channel · prefer email for long threads↵ to send · ↑ for history