Suraj Maurya

Senior Software Engineer

Mumbai, Maharashtra, India3 yrs 9 mos experience

Highly Stable

Key Highlights

Built fault-tolerant systems processing 100M+ events/day.
Reduced model deployment lag from 4 hours to 5 minutes.
Achieved 99.99% uptime in critical data pipelines.

Stackforce AI infers this person is a Backend and Data Engineer specializing in high-throughput, fault-tolerant systems for SaaS applications.

Contact

Skills

Core Skills

Data EngineeringBackend Development

Other Skills

API DevelopmentAWSAWS LambdaAlgorithmsAmazon EC2Amazon ECSAmazon EKSAmazon Relational Database Service (RDS)Amazon S3Back-End Web DevelopmentC (Programming Language)C++CSSContinuous Integration and Continuous Delivery (CI/CD)Data Structures

About

I'm a Backend & Data Engineer with 3+ years of experience designing fault-tolerant, high-throughput systems that scale to 100M+ events/day. I specialize in building real-time ingestion pipelines, self-healing ETLs, and data platforms that power ML, analytics, and product intelligence at scale. 💼 At Shaadi.com, I: Built a modular Feature Store used by multiple ML models, reducing deployment lag from 4 hours to 5 minutes Delivered streaming pipelines (Go + Kafka + DynamoDB) with exactly-once semantics and automated schema evolution Architected Redshift ETL orchestration and feature APIs with 99.99% uptime Reduced failover recovery time by 90% and analytics pipeline incident rate by 70% 🔧 Core Stack: Golang · Kafka · AWS (Lambda, EC2, S3, DynamoDB, Redshift) · Kubernetes · Redis · PostgreSQL · Datadog · Prometheus · CI/CD · CQRS · REST APIs ⚙️ Focus Areas: Event-driven architecture Observability and SLO-driven engineering Schema evolution and ML data infra Production-grade system design with performance SLAs 🚀 Passionate about distributed systems, backend performance, and building platforms that move business metrics.

Experience

Skima ai

Senior Software Engineer

Jul 2025 – Present · 8 mos · Mumbai, Maharashtra, India · On-site

Shaadi.com

2 roles

Software Engineer 3

May 2024 – Jul 2025 · 1 yr 2 mos · Mumbai, Maharashtra, India · Hybrid

🧠 Context:
Spearheaded the architecture and scaling of real-time ingestion systems and ETL pipelines to serve mission-critical backend services at scale.
Achievements:
⚙️ Built Kafka-based ingestion platform (99.99% uptime), reducing failover recovery time by 90%
📊 Engineered self-healing ETL pipelines with schema versioning for 5M+ daily records
⚡ Delivered feature store powering 30+ ML models, reducing model deployment lag from 4 hours to 5 minutes
📉 Introduced observability via Datadog, cutting incident rates by 70% and MTTR by 40%
🧠 Led 10+ design reviews with product, infra, and ML teams, aligning technical delivery to KPIs

KafkaETLDatadogGolangAWSML+3

Software Engineer

Jun 2022 – May 2024 · 1 yr 11 mos · Mumbai, Maharashtra, India · Hybrid

🧠 Context:
Worked across backend and data engineering teams to build scalable systems, event-driven pipelines, and ML infrastructure powering user personalization, insights, and analytics. Played a key role in enabling data-driven decisions through robust, high-throughput architectures.
✅ Key Achievements:
🔹 Feature Platform Engineering
Built a modular, versioned Feature Store with online/offline sync, powering 30+ ML models with sub-150ms latency using Redis caching and LRU eviction.
Reduced model deployment lag from 4 hours to 5 minutes, significantly accelerating model iteration cycles.
Developed real-time counter infra with Redis, rate limiting, and backpressure logic—boosting analytics throughput by 4x.
Designed a unified schema catalog across MySQL, DynamoDB, and Kafka, reducing ML model onboarding time by 50%.
Deployed a Redshift snapshot orchestration system with rollback and freshness tracking, enabling daily retraining and reducing backlog by 70%.
Owned low-latency feature delivery APIs (Golang + API Gateway) with 99.99% SLA uptime.
🔹 Data Engineering & Pipelines
Delivered streaming ingestion pipelines processing 100M+ events/day using Golang, Kafka, and DynamoDB with CQRS and idempotent writes for exactly-once delivery.
Introduced automated schema diffing and evolution tooling, ensuring safe, backward-compatible changes across data pipelines.
Reduced incident rate by 70% through Datadog alerting, auto-remediation workflows, and structured observability tied to SLOs and anomaly detection.
Automated schema validation across batch + stream ETLs, enabling seamless deployments with zero downtime.
🔹 Web Data Extraction Infrastructure
Built a fault-tolerant, multi-threaded web data extraction system in Go with proxy rotation, rate limiting, and dynamic selectors—enabling real-time updates for strategic insights.
Reduced turnaround time for business-critical data by 90%, with zero-downtime experimentation via modular config architecture.