PK_OS · v7.0.26
REC ● LIVE
PK
// signal_open · accepting_transmissions
<identity role="engineer" />

Parv
Khatri_

Full-stack engineer crafting distributed systems, observability pipelines, and LLM-powered tooling that scales to millions. Currently shipping agentic AI at SAP Labs.

> initiate_contact view_archive
parv@portfolio:~
$whoami
>parv.khatri — software engineer @ SAP Labs
$cat focus.json
{
"stack": ["Java", "Spring", "LangChain", "K8s"],
"shipping": "agentic AI + observability",
"scale": "10M+ docs/day, 800+ customers"
}
$uptime
>7+ years · 0 panics
SYS_STATUS
STATUSONLINE
LOCATION37.37°N, 122.04°W
STACKJVM · NODE · PY
UPLINK10Gbps · 0% loss
METRIC_01
0+
Years building
METRIC_02
0+
Enterprise customers
METRIC_03
0M+
Docs / day
METRIC_04
−0%
MTTR cut
JavaSpring BootKafkaKubernetesAWSPostgreSQLReactTypeScriptLangChainPythonDockerTemporalPyTorchRedisMongoDBSAP HANAGrafanaPrometheus

About

Builder of resilient systems.

From single endpoints to globally-distributed pipelines — measured by uptime, latency, and engineer hours saved.

I'm a software engineer with 7+ years shipping APIs, microservices, and AI-driven tooling for enterprise scale. Currently at SAP Labs in Palo Alto, where I lead observability and agentic AI initiatives across 180+ apps.

I led a team awarded Star Team for cloud infrastructure automation, and most recently built an LLM-based log dashboard that cut MTTR by 34%.

M.S. Computer Science from NYU. I write code that other engineers want to inherit.

Star Team Award
TCS — Cloud Infra
Star Learner
Top 1% of 2,000+
NYU Scholarship
Graduate
ISSN 2277-9655
Published
Education
New York University
M.S. Computer Science · GPA 3.78
2020 — 2022
RGPV, Bhopal
B.E. Computer Science
2013 — 2017
Currently
  • Shipping MCP servers for agentic dev tooling
  • Scaling RAG over 10B+ log lines
  • Reading: Designing Data-Intensive Apps

Experience

Where I've shipped.

Three companies, one through-line: take messy systems and make them fast, observable, and boring to operate.

LOG_03

SAP Labs, LLC

Developer

Jan 2022 — Present
Palo Alto, CA
  • Built RESTful APIs for document processing serving 800+ enterprise customers.
  • Shipped microservices processing 10M+ multi-format documents daily.
  • Cut MTTR 34% with an LLM-based log dashboard across 180+ apps.
  • Redesigned background jobs, reducing execution time by 29%.
  • Extended authorization sync service enabling self-onboarding for 800+ customers.
  • Developed MCP services to integrate agentic AI into internal tooling.
JavaSpringKafkaLangChainK8s
LOG_02

SAP America, Inc.

iXp Intern — Full-Stack Developer

Jun 2021 — Dec 2021
Palo Alto, CA
  • Extended SCIM APIs across 12 enterprise apps, saving 3 hrs/week of manual work.
  • Trimmed flaky test failures 20% by optimizing E2E UI automation suites.
  • React and Angular components adopted by 3 internal product teams.
ReactAngularSCIMNode
LOG_01

Tata Consultancy Services

System Engineer

Nov 2017 — Sep 2020
India
  • Scaled REST APIs & UI components for Fortune 500 retail and telecom clients.
  • Led 3-person team automating cloud infra — awarded Star Team.
  • Built 4 microservices automating SAP Basis ops, cutting effort 48%.
  • Top 1% of 2,000+ engineers in TCS technical training (Star Learner Award).
JavaSAPPythonAWS

Selected work

Projects with measurable impact.

Every project ships a number — engineers don't care about features, they care about pages prevented.

Case study

Agentic AI Self-Healing Test Automation

LangChain + Temporal utility that autonomously diagnoses and patches broken CI/CD scripts. Human-in-the-loop LLM workflows repair brittle UI locators and API payloads.

0%
Less maintenance
LangChainTemporalLLMCI/CD
Case study

Observability & Monitoring Dashboard

Kafka pipeline processing 10B+ docs/year across 23 datacenters. Real-time Grafana + Prometheus platform for 10+ microservices with a RAG layer for root-cause analysis.

0B+
Docs/year
KafkaGrafanaPrometheusRAG
Case study

SRE Automation & Infra Monitoring

Automated health checks, cluster management, and downtime recovery across 350+ servers — replacing hours of manual post-release work.

0+
Servers
SREAutomationCloud
Case study

LLM-Powered Log Intelligence

RAG pipeline over 180+ apps' logs. Engineers ask natural-language questions; the model retrieves traces, clusters anomalies, and proposes fixes.

0+
Apps covered
RAGVector DBOpenAIPython

Toolkit

Tech I reach for.

Polyglot by necessity. Opinionated by experience.

MOD_01

Languages

JavaJavaScriptTypeScriptPythonCSQL
MOD_02

Backend

Spring BootExpress.jsDjangoRESTgRPC
MOD_03

Frontend

ReactAngularVueSAP UI5Tailwind
MOD_04

Data

PostgreSQLSAP HANAMongoDBDynamoDBRedisKafka
MOD_05

AI / ML

PyTorchLangChainLangGraphRAGMCPvLLM
MOD_06

Infra

AWSAzureDockerKubernetesTemporal
MOD_07

Testing

JUnitJestKarateSeleniumREST Assured
MOD_08

Observability

GrafanaPrometheusELKOpenTelemetry

Contact

Let's build something.

./transmit --priority=high

Have an interesting problem at the intersection of distributed systems and AI? I'd love to hear about it.

MAIL
khatriparv@gmail.com
VOICE
347-885-3903
LINKEDIN
linkedin.com/in/parvkhatri
GITHUB
github.com/Parv17k
NODE: Sunnyvale.CA · CHANNEL: remote+hybrid

Arcade

Play a quick game.

./arcade --mode=interactive
Neon Flux
Move mouse · Click to detonate
SCORE 0000
▌ rendered via p5.js · click canvas to focus for keyboard input