Headline

Engineering at AngelOne | Ex-MoEngage

About

Senior backend and distributed systems engineer with 8+ years building infrastructure at scale: 150M+ users, 40B+ daily events, 100k+ peak QPS at P99 <20ms. Most recent work: re-architected a financial trading platform's core margin systems (AngelOne) from batch to event-driven, cutting processing time from 24 hours to under 2 seconds and unlocking a Rs.500 Cr+ revenue stream. Owned system design end-to-end — HLD/LLD, API contracts, cross-team rollout across 4 engineering groups. Before that, built and led the ML infrastructure powering MoEngage's recommendation and prediction products: an inference engine serving 50k requests/sec across 100+ enterprise clients, and the Spark/EMR pipelines processing 40B+ events for 150M+ users. Led a 5-engineer team through roadmap, sprint planning, and OKRs; cut infra spend 30% (~$150k/year) via autoscaling and Graviton migration. Core stack: Go, Python, Java, Kafka, Spark, Kubernetes, PostgreSQL, Redis, AWS (EMR, Glue, Lambda, Step Functions), Azure, Airflow, Terraform. Domain depth: event-driven architecture, microservices, stream/batch data pipelines, ML/recommendation systems, LLM context engineering and RAG, performance engineering at scale. B.Tech Computer Science, PES University (CGPA 9.4/10). Published research in ML/cloud computing (IEEE T4E, IEEE CCEM, Procedia Computer Science). Open to Staff Engineer / Tech Lead roles building distributed systems, data platforms, or AI infrastructure at scale.

Location

Bengaluru

AI stack

cursor

opencode

Work experience

Software Engineer 3 at Angel One, Dec 2024 - Present · 1 yr 9 mos, Bengaluru, Karnataka, India, Leading backend engineering on financial infrastructure — Margin Trading Facility (MTF), Margin Pledge, EDIS, and post-trade systems. - Re-architected Loan Against Shares and Mutual Fund Pledge limits from nightly batch to near-real-time event-driven systems, cutting user wait time from 24 hours to 2 seconds and unlocking a Rs.500 Cr+ revenue stream. - Cut container memory 60% and CPU 45% on high-throughput Go microservices via custom connection pooling, sustaining 100k+ peak API requests at P99 <20ms during market hours. - Architected the MTF Shortage Modernization pipeline across 4 cross-functional teams (Product, RMS, PTRMS, SRE), decoupling real-time trading from legacy back-office systems and securing SLA compliance on a Rs.4,000 Cr+ loan book. - Built end-to-end CI/CD for AWS Glue, Lambda, and Step Functions, enabling one-click disaster recovery failover across regions. - Cut MTTR 40% and non-peak alert noise 70% by redesigning incident response, monitoring dashboards, and runbooks. - Owned end-to-end delivery of the Price Craft Engine for Plan Management; authored HLD/LLD and API contracts that became the cross-team design standard at AngelOne. Stack: Go, AWS (Glue, Lambda, Step Functions), PostgreSQL, event-driven architecture.

Lead Software Engineer at MoEngage, Mar 2021 - Nov 2024 · 3 yrs 9 mos, Bengaluru, Karnataka, India, Led the Data Science engineering team building ML infrastructure and recommendation systems for MoEngage's enterprise marketing platform. - Built and shipped an AI recommendation engine on AWS serving 50k inferences/sec (real-time + batch) for 100+ enterprise clients, lifting campaign CTR/CVR 4% platform-wide. - Led a 5-engineer team — owned roadmap, sprint planning, and OKRs with Product/EM; mentored 2 engineers to mid-level promotion within 18 months. - Scaled Spark/EMR pipelines to process 40B+ events across 150M+ users through query tuning and cluster optimization, cutting customer onboarding time 25%. - Reduced infra spend 30% (~$150k/year) via AWS autoscaling tuning, Graviton migration, and a rule engine that right-sized infra per customer. - Made all microservices cloud-agnostic, refactoring Spark workloads from EMR to Kubernetes to enable Azure market entry. - Shipped MoEngage Predictions (Best Time to Send, Preferred Channel, Predictive Segments) end-to-end — adopted by 60%+ of enterprise clients. - Architected the AutoReports feature, automating 1M+ daily scheduled dashboard-chart emails to customers and reducing manual analytics overhead to zero. Stack: Python, Spark, EMR, Kubernetes, Airflow, AWS, Azure, Kafka.

Software Engineer at Instamojo, Jul 2019 - Feb 2021 · 1 yr 8 mos, Greater Bengaluru Area, - Designed a reconciliation microservice and Kafka-based ingestion pipeline processing 10M+ daily transactions, improving GMV calculation accuracy from 80% to 99%. - Cut payment failure rates 15% by resolving reliability bottlenecks in the core payments stack. - Built a gateway transaction parsing engine from scratch — settlement file parsing, transfer service, system-level alerting — enabling automated financial reconciliation at scale. Stack: AWS, Kafka, Python, payments infrastructure.

Software Development Intern at Instamojo, Jan 2019 - Jul 2019 · 7 mos, Bangalore

Machine Learning / AI Intern at Aiaioo Labs, Jun 2018 - Jun 2018 · 1 mo, Bengaluru, Karnataka, India, - Built a Named Entity Recognition model on the CoNLL-2003 benchmark dataset for core NLP tasks. Stack: NLP, Machine Learning, Named Entity Recognition.

Research Intern at Center for Cloud Computing and Big Data, PES University, Sep 2017 - Nov 2018 · 1 yr 3 mos, PES University, Bengaluru, - Built a novel mispronunciation classifier for spoken-language assessment — analyzed pronunciation quality and localized errors to specific phonetic segments using speech recognition and deep learning. - Work led to 2 published research papers and an IEEE Best Innovative Project Award (Proof of Concept). Stack: Speech Recognition, Deep Learning, Machine Learning.

Education

Bachelor’s Degree, Computer Science, PES University