connection established

~/shailesh — bashstatus: 200 ok
$whoami
Shailesh Singh
$role
AI Engineer · Real-time AI Systems
$company
Zykrr · Sep 2025 — Present
$location
Gurgaon, Haryana, India
$stack | head
Gen AI · ClickHouse · K8s · LiveKit · LangChain
$

Building real-time AI: distributed CX pipelines, low-latency voice agents, and LLM-driven analytics at production scale.

resume.pdf
·press tab to traverse · pid 2026
~/system-render.glsl● live
// loading geometry...
fps: ~60geom: icosahedron@1.4r{drift, bob, rot}
$ cat /var/log/zykrr/impact.json
metric_00
34.56M+
events/day
Kafka · Postgres → ClickHouse
metric_01
STT → LLM → TTS
voice pipelines
low-latency on LiveKit
metric_02
K8s
distributed scaling
Kafka consumers · HA
metric_03
DLQ
fault tolerance
recovery · production debugging
$ git log --author="Shailesh" --oneline

~/work_history

// commits in reverse chronological order

end of log · 2 commits shown
$ cat about.md

~/about_me

~/identity.pngimage · 100%
Portrait of Shailesh Singh
● recordingGurgaon
phone
+91-6201237193
email
ss.forcoding@gmail.com

AI engineer with a systems mindset. I build real-time AI infrastructure — from streaming pipelines to LLM analytics to conversational voice agents — and I'm happiest when latency, correctness, and resilience all have to hold at once.

I'm an AI Engineer at Zykrr where I design real-time systems that turn live customer feedback into structured insight. My day-to-day spans Apache Kafka pipelines moving 34.56M+ events/day from PostgreSQL to ClickHouse, LLM-powered analytics (text mining, issue detection, summarization), AI-driven survey builders, and low-latency STT → LLM → TTS voice agents on LiveKit. I lean toward fault-tolerant design — Kubernetes, DLQs, recovery — and care about latency, accuracy, and what happens when production breaks at 3am.

$ tags --grep .
[ai/ml]
LLMs · RAG · LangChain · agents
[real-time]
Kafka · ClickHouse · LiveKit
[distributed]
Kubernetes · MSK · async
[production]
DLQ · debugging · observability
$ education --institute IET-Lucknow --verbose

~/education

Institute of Engineering and Technology, Lucknow

B.Tech, Computer Science and Engineering

Oct 2021 — Jun 2025Lucknow, U.P., India
cgpa
8.04 / 10
$ ls -la ~/.stack

~/stack

// hover a category to highlight its nodes

$list --languages5 entries
Languages
PythonC/C++JavaScriptTypeScriptSQL
$pip list | grep ai/ml10 entries
AI/ML & GenAI
LLMsRAG PipelinesPrompt EngineeringLangChainCrewAILangGraphHugging FacePyTorchConversational AIText Analytics
$ls ~/backend7 entries
Backend & Systems
FastAPINode.jsDistributed SystemsMicroservicesREST APIsAsync ProcessingApache Kafka
$kubectl get pods --all8 entries
DevOps & Cloud
DockerKubernetesAWS MSKAWS IAMAzureGCPLiveKitCI/CD
$psql -c '\dt'6 entries
Databases
ClickHousePostgreSQLMongoDBChromaDBNeo4jVector DBs
$echo $STRENGTHS5 entries
Core Strengths
Production DebuggingPerformance OptimizationReal-time ProcessingFault-tolerant SystemsScalable System Design
~/stack-graph.tsx41 nodes
// indexing nodes...
// hover a panel on the left
$ ls ~/projects --featured

~/projects

stratagi/README.md● featured
StratAGI
● featured

# StratAGI

> Autonomous Multi-Agent Business Strategy Platform

A multi-agent AI system using CrewAI for autonomous research and strategy generation workflows. Reduced redundant API calls by 30% through optimized orchestration and shared memory design. Backed by FastAPI services and shipped in containers.

30% fewer API calls via shared memory orchestration
// requirements.txt
PythonCrewAILangChainFastAPIDockerStreamlit
wealthwise/README.md● featured
WealthWise
● featured

# WealthWise

> Full-Stack AI Finance Platform

A full-stack AI finance platform for automated expense tracking, budgeting, and financial insights. Implements LLM-based receipt parsing to extract structured data from unstructured inputs, with real-time recurring transactions, alerts, and reporting workflows.

Receipt → structured JSON via LLM parsing
// requirements.txt
Next.jsTailwind CSSPrismaClerkGemini AIVercel
$ ssh shailesh@portfolio

~/contact

// open a session — i'll reply within 24h

contact.jsonreadonly
{
  "name":     "Shailesh Singh",
  "role":     "AI Engineer · Real-time AI Systems",
  "email":    "ss.forcoding@gmail.com",
  "phone":    "+91-6201237193",
  "location": "Gurgaon, Haryana, India",
  "timezone": "Asia/Kolkata (IST)",
  "status":   "open to opportunities"
}
Gurgaon, Haryana, India
session active · /dev/pts/0
>*
>*
>*
>
built with next.js · framer-motion · three.js
© 2026 Shailesh Singh · all rights reserved