01

About

I'm a software engineer with 6+ years of experience building production systems that have to work — at scale, under pressure, and over time. My current focus is distributed systems, backend infrastructure, and API design.

Right now I'm a Senior Product Engineer at KhiladiPro, where I scaled assessment throughput from 5 to 10,000/day using a Flyte ML pipeline on GCP, cutting report turnaround from 5–7 days to 20 minutes. I've owned systems end to end — database schema to Kubernetes deployment.

I care about the craft: clean API contracts, resilient failure modes, readable code. I've been a founding member at an early-stage startup, led microservice architecture, and worked across the full stack when the situation called for it.

Outside of engineering — open water scuba diver 🤿 and guitarist 🎸. Time offline makes for better systems thinking.

6+ Years exp
5 Companies
10K Assessments/day
10TB Data owned
02

Experience

KhiladiPro Senior Product Engineer
Mar 2024 – Present
  • Architected the assessment processing pipeline (Go, Flyte, GCS) — scaled drill processing from 5 to 10,000/day (2,000x) and eliminated 90% of manual ML-execution effort.
  • Led the assessments-domain backend redesign (Go, PostgreSQL, GCS) — collapsed report SLA from 5–7 days to 20 minutes and unified 4 customer surfaces on a single canonical schema.
  • Designed a resumable chunked-upload protocol ingesting 3 TB/day of athlete video into the ML pipeline (Go, GCS) — zero data loss from zero-network venues.
GoFlyteGCSPostgreSQLGCPKubernetes
Sapper Senior Software Engineer
May 2023 – Mar 2024
  • Architected the Securities & Custodian Account service (Spring Boot, JPA, Keycloak) for Genpact Cora LiveWealth — aggregated 55K+ accounts across 300+ custodians and cut quarterly reporting from 45 to 5 days (75% efficiency gain).
  • Productized an Open Data Discovery fork into Maitrx, Genpact's enterprise data catalog (Spring Boot, OpenAPI 3, React), owning the search subsystem — federated metadata across 7+ microservices and reduced data-asset discovery from hours to sub-second.
  • Built the Flowable + Kafka + HBase integration powering Cora LiveWealth's OCR document-extraction pipeline — collapsed custodial-statement reporting from 30 days to real-time.
JavaSpring BootJPAKeycloakOpenAPI 3ReactKafkaHBaseFlowable
Tao Founding Member
Mar 2022 – May 2023
  • Built order management and real-time leaderboard services from scratch (Spring Boot, MongoDB, Redis) — held P95 latency under 100ms at 1,000 DAU across the store and competition flows.
  • Led a full Android rewrite (Kotlin, MVVM, Jetpack Compose) in 1 month — lifted crash-free sessions from 70% to 99.8% and raised user retention 20%.
  • Set up the engineering function as Tao's first hire — instituted CI/CD and code review for a 5-engineer team and mentored 3 ICs, cutting production defects 20%.
Spring BootMongoDBRedisKotlinJetpack ComposeMVVM
GetMega Software Engineer
Sep 2020 – Mar 2022
  • Owned Android delivery for the GetMega gaming product (Kotlin, Coroutines, Jetpack Compose) — sustained 20–30% MoM revenue growth across 100K MAU.
  • Hardened the payment integration layer (REST, gRPC, third-party payment SDKs) — cut transaction failures 30% across the financial flow.
  • Cut Android cold start 20% and lifted crash-free sessions to 98% via startup profiling and memory tuning.
KotlinCoroutinesJetpack ComposeAndroidgRPCREST
Novopay Software Engineer
Jul 2019 – Sep 2020
  • Redesigned the white-label Android app templating architecture (Java, MVVM, Dagger) across 3 banking products — cut client onboarding from 2–4 weeks to 1 day.
  • Integrated payment flows handling INR 18–22L/month profit — improved reconciliation accuracy 15% and cut transaction errors 15%.
  • Automated the multi-product Android release pipeline — cut deployment time 50%.
JavaMVVMDaggerAndroidPaymentsCI/CD
Amazon SDE Intern — UPI Payments Team
Jun 2018 – Aug 2018
  • Built an automated notification workflow on the UPI Payments team (AWS Lambda, SNS, DynamoDB) — compressed operational resolution from 4–6 days to 5 minutes and cut on-call load 10–15% across 2 teams.
AWS LambdaAWS SNSDynamoDBJava
03

Projects

Rocky — AI Gym Coach

A mobile fitness app that acts as an AI coach, not a chatbot. Rocky runs a perceive–reason–act loop nightly over your training history and delivers a fully-planned session as a morning briefing — the thinking is done by the time you open the app.

Adapts plans in seconds when you flag fatigue, soreness, or a tight schedule. Parses natural-language entries like "bench 3x5 at 82.5kg, squats 5×5 at 100" and backdates them with a review step. Two-slider post-workout debrief, a coaching notebook you can read, and zero ads, trackers, or social feed.

AI Agent LLM iOS Android Mobile
Visit Project

Kitchen Nightmare

A cost-efficient household grocery management and recipe suggestion assistant that leverages multiple LLMs — Claude, open-source models — to deliver flexible, budget-aware meal planning.

Modular architecture lets you swap LLM providers to optimize for cost and quality. Tracks what's in your kitchen, suggests recipes from available ingredients, and helps plan grocery runs — all without locking you into a single AI provider.

Claude Open-source LLMs Multi-LLM Python
Visit Project
View more on GitHub
04

Skills

Languages

GoJavaKotlinTypeScriptSQL

Backend & APIs

Spring BootgRPCRESTOpenAPI 3Flowable

Data & Messaging

KafkaHBasePostgreSQLElasticsearchRedisMongoDBDynamoDBFirestore

Cloud & Infrastructure

GCPAWSDockerKubernetesGitHub ActionsCloud Build

ML & AI

FlyteMediaPipeTFLiteGemini

Mobile

AndroidJetpack ComposeMVVMKMMReact Native
05

Education

Indian Institute of Technology (IIT) Mandi

B.Tech in Computer Science & Engineering

Mandi, Himachal Pradesh, India

2015 – 2019
What's next?

Let's build something together.

Whether you have a project in mind, want to talk distributed systems, or just say hi — my inbox is always open.

Say Hello