2020–2026
ConvertCart
Backend Engineer, Platform Ownership — Data ingestion, warehouse consolidation, multi-channel personalisation
Staff Backend Engineer
I build the backend systems behind data-heavy products.
13 years backend systems · 6 years platform ownership Bengaluru, India · Remote
200+
stores
sub-minute
freshness
98%+
uptime
Selected Systems
01 DATA PLATFORM
Rebuilt the platform serving 200+ merchants. Sub-minute freshness for webhooks, up to 6 hours for custom integrations.
02 CUSTOMER SEGMENTATION
Infrastructure behind 1,000+ behavioural segments. Enabled non-technical staff to build audiences without filing tickets.
03 RECOMMENDATIONS
Multi-region recommendation serving. 80%+ adoption on hardest tier; 200ms p99 latency across five regions.
04 MICROSERVICES FLEET
57+ coordinated microservices. Built CI/CD automation and multi-service observability across the fleet.
Specialties: Backend architecture · Data platforms · Distributed systems · Production migrations · Technical leadership
Five years owning a data platform end to end — ingestion, modelling, serving, and the cost of running it. Two CDC paths, a warehouse consolidation, a multi-tenant landing store, and a reporting migration driven by the shape of a billing model.
By the numbers
sub-minute
Data freshness
webhooks; up to 6h custom
200ms
p99 latency
serving
4-person
Team size
3-month migration
Selected work
The data platform
Rebuilt how customer and order data flows through the company — from data arriving days late to under a minute, across 100+ merchant stores on five e-commerce platforms.
days → sub-minute98%+ uptime200+ stores
Customer segmentation
Built the system that lets a non-technical Customer Success rep build audiences without writing a line of code or filing an engineering ticket. The hardest thing I built.
4 source systems1 query1,000+ segments
Recommendation blocks
Built the company's product recommendation feature from nothing, designed so store owners could trust it enough to turn it on. 80%+ adoption on the hardest tier.
80%+ adoption5 regions200ms p99
Also shipped
Microservice fleet at ConvertCart
Recommendation ranking, email delivery, analytics ingestion, webhook distribution — split from the monolith into coordinated services. Owned deployment, configuration, and observability across the fleet. Built CI/CD automation and multi-service observability across 57+ microservices.
MySQL optimisation at Scientific Games
Lottery reporting queries running 45 minutes daily on an un-indexed 20M-row schema. Analysed and restructured queries, designed covering indexes, partitioned tables by date. Reduced the same workload from 45 minutes to 3.
Email infrastructure at Tenlegs
Built the data infrastructure supporting millions of emails/day — recommendation feeds and transactional data pipelines providing all data to the email delivery system. Worked with the email team to ensure data freshness and reliability for high-volume sending.
Activity
checked just now
Independent products
affairsmap.com
Current-affairs study platform for UPSC, SSC, Banking, RBI Grade B and Defence exam aspirants.
In development
Fruggy
Grocery shopping planner for Indian FMCG shoppers. Flutter app with a dataset of 1,500+ FMCG items.
1,000+ downloads
HitReplAI
AI-powered social media reply generator that learns and mimics the user's writing style.
Live on Chrome Web Store
en-bus
Tamil Nadu bus routes and timings finder. Real-time route search across TNSTC, SETC, and MTC buses.
Live · 93K+ tripsStack
Node.js
TypeScript
MySQL
Debezium
Pub/Sub
MongoDB
BigQuery
ClickHouse
Redis
GKE
Principles
Reversibility
Refuse to normalize at ingestion where the decision is irreversible. Normalize in the warehouse where a wrong model is a redefinition, not a migration.
Match storage to question
No single database serves all workloads. Point lookups need one shape, aggregation needs another. Accept the operational cost of multiple shapes matched to their use.
Operational simplicity
Abstractions cost clarity under pressure. Ship what's straightforward to operate and debug. Only abstract where the operational burden actually drives the design.
Skills
Deep
Data pipeline design — end-to-end ingestion, CDC, multi-layer servingStreaming & CDC — Debezium fan-out to Pub/SubBigQuery & analytics — Datastream replication, materialized viewsClickHouse — schema design, partitioning, ORDER BY tuningSystem design — microservices, distributed tracing, reliabilityE-commerce platforms — Shopify, BigCommerce, WooCommerce, Magento
Working knowledge
KafkaApache BeamKubernetesGCP infrastructureReactReact NativeRedisPostgreSQL
Leadership & Scale
Team hiring & evaluationArchitecture ownershipProduction incident responseLarge-scale migrationsCross-functional partnerships
Background
13 years building backend systems and data infrastructure. Started building for the web in 2010. Spent nearly six years at ConvertCart owning a data platform end to end and leading a team of engineers to rebuild it. Built segmentation and recommendation engines on top. Founded and shipped independent products. Experienced hiring and evaluating backend/data engineers; comfortable in both IC and multiplier roles.
A1
A2
2018–2019
Friday Media Group
Backend Engineer, Founding Team — Marketplace APIs, content ingestion, data pipeline
A3
2014–2018
Scientific Games
Backend Engineer — Lottery reporting, data optimization, microservices
A4
2012–2014
Tenlegs
Full-stack Engineer — Email systems, logistics APIs, operational tools
A5
2010–2012
ISPG Technologies
Junior Developer — E-commerce backend, MySQL optimization
Open to high-ownership roles