Open to senior backend / AI-platform roles

Garvit Jindal

I build backend systems and LLM pipelines that hold up at scale.

● Currently at Moody's · Previously SAP Labs · Bengaluru, India

Get in touch Résumé ↗ See the work ↓

garvit@portfolio — zsh

❯ whoami

Backend · Cloud · LLM Engineer

❯ cat focus.txt

system design · automation · developer productivity · production LLM systems

❯ uptime

3+ yrs · SAP Labs → Moody's · healthy

❯ ls projects/

llm-observability data-pipelines TIP100 RideFusion

❯ git log --oneline -1

3395386 feat: add config endpoint

❯

scroll

// system.metrics github.com/GarvitJ09 ↗

3 +

Years engineering

500 +

SAP services automated

10 +

Apps migrated to cloud

75 %

Manual QA effort removed

40 %

Production downtime cut

90 %

QA test time cut at SAP

// recent.activity

Garvit Jindal's GitHub contribution graph — open-source activity

// where.i.do.my.best.work

Three areas, deep

Backend & Distributed Systems

Java/Spring Boot microservices and Python pipelines handling millions of records with sub-second latency. Strong on API design, async event-driven architectures, and migrating legacy services to cloud-native platforms.

Java
Spring Boot
Python
Node.js
REST
Microservices
Async

AI & LLM Engineering

Production-grade LLM systems: FastAPI + LangChain with SSE streaming, model-agnostic provider switching, and an observability layer that tracks cost, tokens, and time-to-first-token per model — with a Redis hot tier and MongoDB for low-latency conversational memory.

Python
FastAPI
LangChain
OpenAI
Gemini
SSE
Redis
MongoDB

Cloud & Data Infrastructure

ETL pipelines and cloud-native services on AWS and SAP BTP. Currently building AWS Glue + Python Lambda pipelines that process and validate large-scale financial data, with operational tooling for the teams who depend on it.

AWS Glue
Lambda
S3
SAP BTP
Cloud Foundry
MongoDB
Docker

// selected.work

Systems I've shipped

Each one: the problem, what I shipped, and the number that moved.

Smart India Hackathon 2022 · National winner ↗

TIP 100

problem: Sensitive intelligence tip-offs needed truly anonymous reporting.

Secure, encrypted tip-off submission system. National winner among thousands of teams.

🏆 National winner

Web
Security

Side project ↗

RideFusion

problem: Riders juggle multiple apps to compare fares and ETAs.

Aggregator that books and tracks rides across providers from one interface, comparing fares and ETAs in real time.

Multi-provider

Mobile
APIs
Integration

// experience.log

Where I've worked

Moody's Corporation

Dec 2025 — Present

Software Engineer (SDE-2)

› Building event-driven data pipelines on AWS (Glue, Lambda, S3) that process and validate large-scale financial datasets.
› Designing scalable ETL and serverless data processing for structured and semi-structured financial data.
› Building internal operational tooling so data teams can monitor pipeline output and act on it.

Python
AWS Glue
Lambda
S3
ETL

SAP Labs

Jul 2023 — Dec 2025

Developer Associate (SDE-1)

› Built a regression scoping tool that cut QA testing time by 90% and manual effort by 75%, adopted across enterprise testing workflows.
› Automated a service classification system in Java/Spring Boot, improving infrastructure efficiency across 500+ SAP services.
› Migrated 10+ legacy applications to SAP BTP / Cloud Foundry, reducing production downtime by 40%.
› Designed backend APIs connecting cloud-based SAP tools with internal systems.

Java
Spring Boot
SAP BTP
Cloud Foundry
HANA

SAP Labs

Feb 2023 — Jul 2023

Developer Associate Intern

› Onboarded onto the SAP BTP and HANA cloud stack; contributed to early-stage backend service development and internal tooling.

Java
SAP BTP
HANA

Axact Studios

Jun 2022 — Jul 2022

Web Development Intern

› Front-end development and UI/UX improvements for client-facing web projects.

MongoDB
Express
React
Node.js
Firebase

// writing --recent

Thinking out loud

All posts on Medium ↗

RAG · agents

RAG Isn't Dying. It's Moving Out of Your Pipeline and Into Your Agent's Hands.

Why retrieval belongs inside the agent's tool-use loop instead of a pre-baked pipeline — and what that changes about context-window design.

Medium · 12 min read ↗

memory · LLM apps

Your AI Doesn't Remember You. Here's the Pattern That Changes That.

A session-memory pattern for stateless LLM apps — Redis hot tier, Mongo persistence, and the ergonomics that make a conversation feel continuous.

Medium · 8 min read ↗

// about

I'm a software engineer with 3+ years in backend and cloud data engineering — currently at Moody's, previously SAP Labs. I build scalable backend systems and cloud-native architectures, specializing in Java (Spring Boot), Node.js, and Python — microservices that handle millions of records with sub-second latency.

Recently I've been deepening my AI engineering work: production LLM orchestration, observability, and low-latency conversational systems. Winner of Smart India Hackathon 2022. B.Tech in Computer Science from JIIT.

// recognition

Rising Star Award — SAP Labs Sept 2025
Publication — QMate Regression Testing with Advanced Scoping Aug 2024
Smart India Hackathon — National Winner Aug 2022

// now Bengaluru · now

Building event-driven data pipelines on AWS at Moody's.
Writing in public on LinkedIn & Medium.
Aiming to shape how people around the world find their next opportunity.

// education

Jaypee Institute of Information Technology

B.Tech, Computer Science · 2019 — 2023

Focus on software development, data structures, and database systems. Active in technical competitions and hackathons.

// contact --open

Let's build something

Hiring for a senior backend / AI-platform role, or want to talk through a system you're shipping? I read every message.

garvit992001@gmail.com ↗

LinkedIn ↗ GitHub ↗ Twitter ↗ LeetCode ↗

new-message --to garvit