KAP.dev

// Melbourne, Australia · open to backend & cloud roles

Kaviraj Ananthapuri
Palanivel

Full-stack engineer, cloud-native — building with AI, grounded in real engineering. React/Next on the front, NestJS & FastAPI microservices behind, shipped on AWS & GCP — plus data engineering and a deepening focus on AI.

See the workRésumé ↗
scroll to explore

01 — About

Full-stack & cloud-native — and going deeper every week.

Full-stack and cloud-native engineer finishing a Master of IT at Monash (GPA 3.875/4.0). I build end-to-end — React/Next on the front, NestJS and FastAPI microservices on the back, deployed across AWS and GCP with Docker, Kubernetes, and IaC. Lately I'm going deep on AI (benchmarking models across hardware, learning the ML foundations) and on the craft itself: system design, low-level design, and the kind of problem-solving that AI tools can't shortcut.

I care less about chasing every new framework and more about engineering judgement — clean architecture, efficient code, and solving problems that genuinely need experience. AI makes me faster; sharpening my fundamentals makes me valuable.

Master of Information Technology

Monash University, Melbourne

GPA 3.875/4.0

Expected 2026

Bachelor of Computer Applications

Amrita Vishwa Vidyapeetham

CGPA 9.28/10.0

2020 – 2024

Exchange Semester, Informatics Engineering

Universitat Politècnica de València

2023

0+

LeetCode solved

0.000/4.0

Master's GPA

0.00/10

Bachelor's CGPA

Currently building & learning

Benchmarking AI models across hardware

Measuring how different models perform on different hardware — throughput, latency, efficiency.

ML foundations, properly

Supervised vs unsupervised learning, classification, and how models actually work under the hood.

Engineering craft

System design, low-level design, and writing efficient code — the depth AI tools can't replace.

Sharper problem-solving

Leaning on AI for speed while building the judgement to solve problems that need real engineering.

02 — Selected work

Things I've built.

From on-device ML and Kubernetes inference services to event-driven microservices. Hover a card.

AI Model Benchmarking

in progress

Performance of AI models across hardware

2026
  • Benchmarking how different AI/ML models perform across different hardware — throughput, latency, and efficiency tradeoffs.
  • Part of going deeper on the systems side of AI: how models actually run, not just how to call them.
PythonMLBenchmarkingHardware Profiling

WildQuest

Gamified Wildlife Discovery App

2026
  • Android app for discovering, photographing, and logging real animals nearby, with on-device species recognition via a TensorFlow Lite model (MobileNetV2, transfer learning) and a custom FastAPI backend serving location-specific animal data.
  • Location-aware quests on Google Maps (custom markers, polylines, live tracking), background step tracking via WorkManager, Firebase auth, and offline persistence with Room.
KotlinComposeFastAPITensorFlow LiteFirebase

CloudEco

Marine Plastic Detection Service

2026
  • Container-orchestrated ML inference service (FastAPI + YOLOv8m) detecting marine plastic via REST APIs on a highly available GCP Kubernetes cluster provisioned with Terraform and Ansible.
  • Sped up inference with OpenVINO INT8 and load-tested using Locust.
FastAPIKubernetesTerraformAnsibleGCPYOLOv8

Advanced Databases — Data Engineering & EDA

Cleaning, transforming, and finding insight

2025
  • End-to-end data work: cleaning, processing, and transforming raw datasets through the full data-engineering pipeline.
  • Post-cleaning exploratory data analysis to surface valuable, decision-ready insights for the client.
PythonSQLEDAETLAnalytics

byteSized

Telehealth Platform — MedHack 2026

2026
  • Designed and pitched a telehealth platform with AI-based diet tracking and patient assistants.
  • Led product strategy, system design, and high-fidelity prototyping through the pre-implementation phase.
System DesignAI IntegrationPrototyping

PTV Bus Transport Analysis

Transport Equity, Greater Melbourne

2025
  • ogr2ogr ETL pipeline ingesting 10,290+ mesh-block records into PostGIS.
  • Assessed transport equity via a service-to-population model, optimised using GIST spatial indexes.
SQLPostGISQGISDocker

Access Key & Token Services

Web3 Microservices

2024
  • Two NestJS microservices for access-key management and Web3 token retrieval over RabbitMQ.
  • Key generation, rate limiting, and TTL validation backed by Prisma/PostgreSQL with a request guard enforcing access control.
NestJSRabbitMQPrismaPostgreSQLDocker

03 — Toolbox

The stack I reach for.

Languages

JavaPythonTypeScriptJavaScriptKotlinC++C#SQL

Frontend

ReactNext.jsTailwind CSSBootstrapJetpack Compose

Backend & Architecture

NestJSFastAPINode.jsRabbitMQMicroservicesSystem DesignLow-Level DesignDSA

Cloud & DevOps

AWSGCPDockerKubernetesTerraformAnsibleCI/CDGitAgile/Scrum

Data Engineering & Analytics

ETLEDAData CleaningTransformationPostgreSQLMongoDBPrismaPostGISInsights & Analysis

AI / ML (actively growing)

Model BenchmarkingAI IntegrationSupervised / UnsupervisedClassificationTensorFlow LiteOpenVINO

Testing & Quality

JUnitMockitoSonarQubeLocustTDDClean Code

04 — Experience

Where I've shipped.

Associate Software Engineer Trainee · NeoITO

08/2024 – 11/2024

Trivandrum, India

  • Built scalable microservices with NestJS and RabbitMQ, implementing event-driven communication between distributed services to improve throughput and reduce API bottlenecks.
  • Applied Clean Code principles and Test-Driven Development within an Agile team to improve reliability and maintainability of the codebase.

Teaching & Leadership

Mentoring · clubs · TA (upcoming)

  • Ran student clubs during my undergrad — organising, mentoring, and teaching peers.
  • In progress: Teaching Assistant role for an upcoming Monash unit.
  • I genuinely enjoy explaining hard things simply — teaching is where I do my best thinking.

Certifications

  • AWS Certified Cloud Practitioner
  • AWS Academy Graduate
  • Python for Data Science, AI & Development

Awards

  • Highest Grade Commendation (Java, Monash FIT)
  • Excellence in Teamwork
  • Academic Excellence

05 — Arcade

Stay a while and play.

Four tiny games built around tech and computer science. No high score table — just bragging rights.

Question 1/5Score 0

Which data structure works on a Last-In, First-Out (LIFO) basis?

06 — Contact

Let's build something reliable.

Open to full-stack, cloud, platform, and AI/ML roles. The fastest way to get a feel for my work? Ask my AI host — or just email me.