Headline

SDE-II Platform Engineer @MoneyView | Kubernetes (EKS) | AWS | Internal Developer Platform (IDP) | Terraform | Observability | CI/CD | DevOps | Musician

About

I spend my days making life easier for other engineers — building internal platforms, cutting cloud cost, and untangling the kind of problems that quietly slow teams down. I'm at my best when the fix isn't just a better tool, but a better way of working around it. My work lives mostly in DevOps and platform engineering — self-service infrastructure, observability, cost, and the occasional security rabbit hole. I care less about which tool wins the debate and more about whether the team around it is actually unblocked. Music is the other half of my brain. Engineering rewards you for being clever; music only rewards you for being true. Open to interesting roles at companies with a mission.

Location

Mumbai

AI stack

claude code

claude

cursor

ollama

openai

replit

Work experience

SDE at Moneyview, Feb 2023 - Present · 3 yrs 9 mos, Bengaluru, Karnataka, India, Platform Engineer who builds the internal systems other engineers depend on — turning manual infra work into self-service platforms. 3 years at Moneyview on AWS + Kubernetes. ━━ Scope ━━ (1) Owned infra provisioning + production support across 15+ services over multiple AWS accounts and regions — SQS/SNS/Lambda/S3/Firehose, staging/DR setup, cross-account networking. ━━ Monitoring & Observability ━━ (1) Built Arirang — a Java-based in-house PagerDuty replacement with hierarchical on-call escalation. Alerts reach the engineer + PM in under 5 minutes. (2) Migrated the log pipeline from Filebeat to Vector after root-causing Filebeat's ALB incompatibility — ended production data loss. (3) Built a JVM diagnostics toolkit auto-capturing heap/thread dumps + GC logs to Datadog during incidents. ━━ Infrastructure & Security ━━ (1) Architected InfraPilot — self-service AWS provisioning with auto-generated Terragrunt configs, approval workflows, and audit trails. Infra requests cut from days to ~11 minutes. (2) Led EC2 → EKS migration with a fill-in-the-blanks onboarding utility — containerization cut from 3–4 hours to ~22 minutes (~88%). (3) Built SpotD — a proactive spot-instance manager with on-demand fallback. 60% staging compute savings, zero downtime. (4) Drove cost optimization: VPC Endpoints, Aurora Serverless, RDS Reserved Instances (~40%), S3 lifecycle, Slack cost-anomaly alerts. (5) Ran 2 company-wide DR drills with live traffic in ap-south-2, 24 hours each. (6) Contributed to org-wide end-to-end mTLS rollout (zero-trust service comms). (7) Shipping now: Karpenter, KEDA, Kyverno, Velero/AWS Backup, Cilium network policies. ━━ CI/CD ━━ (1) Introduced Jenkins Shared Library pattern across teams. (2) Offloaded Jenkins jobs to AWS CodeBuild — eliminated Jenkins downtime. (3) Co-architected a self-service release process so devs ship their own releases. Stack: Kubernetes (EKS), AWS, Terraform/Terragrunt, Datadog, Jenkins, CodeBuild, Vector, Java , Python , ELK.

SDE - Intern at Jio, Mar 2021 - Jul 2021 · 5 mos, Hyderabad, Telangana, India, Contributed to a sports analytics CV pipeline focused on cricket, spanning dataset creation, annotation tooling, and real-time object detection research. (1) Implemented Multiple Object Tracking (MOT) for cricket ball trajectories, handling fast-motion and occlusion challenges specific to sports footage (2) Built and curated a large-scale cricket shot dataset covering batsmen and bowler actions — annotated and augmented over 1M image frames used for downstream model training (3) Performed object segmentation using automated tools including ZVOS (AGNN) alongside manual annotation workflows in CVAT to generate high-quality training masks (4) Worked on frame change detection to identify key events and segment long-form cricket video into analyzable clips (5) Evaluated and compared real-time object detection architectures — YOLOv4, Faster R-CNN, SSD — benchmarking accuracy/latency trade-offs for the cricket use case (6) Wrote Python utilities to process and merge CSV annotation outputs, standardizing dataset formats across the pipeline Tech: Python, OpenCV, YOLOv4, Faster R-CNN, SSD, CVAT, ZVOS (AGNN)

Machine Learning Intern at Tata Group, Apr 2020 - May 2020 · 2 mos, Mumbai, Maharashtra, India, Applied classical ML techniques to analyze customer payment and loan behavior, supporting data-driven decisions in the micro-finance business line. (1) Built regression models (Linear and Logistic Regression) to analyze borrower payment patterns and loan behavior, identifying signals correlated with repayment outcomes (2) Performed exploratory data analysis and feature engineering on micro-finance datasets to surface the variables most predictive of customer behavior (3) Tuned and adapted model parameters to fit the specific characteristics of the micro-finance use case, iterating on feature selection and model configuration to improve fit Tech: Python, scikit-learn, pandas, NumPy

Education

Bachelor of Technology - BTech, Electrical, Electronics and Communications Engineering, Maharashtra Institute of Technology

Nerul, Navi Mumbai, Delhi Public School - India

Public links

github: https://github.com/Akshath11

medium: https://medium.com/@Akshathagrawal

leetcode: https://leetcode.com/aki1108