Based in Sydney, Australia · AEST

Guru
Prasad Raju

I'm a Cloud Automation Engineer

I build smart cloud ecosystems that simplify complexity, reduce friction, and scale automatically — from AWS architectures to AI-powered automation frameworks.

View Projects
~/projects
guru@aws ~ $ aws lambda list-functions --query 'Functions[*].FunctionName' [ "auto-scaler", "cost-optimiser", "llm-inference-proxy" ]   guru@aws ~ $ python run_llm.py --model llama3 --local # Loading model weights (4-bit quantised)... ✓ Model ready in 8.3s · 3.9 GB RAM   guru@aws ~ $ pytest tests/ --headless -n auto # Running 248 UI automation tests... ✓ 248 passed · 0 failed · 12.4s   # Automation rate: 94% · Flakiness: <2%
What I bring

Core Expertise

Three pillars that power the work I do — cloud, automation, and delivery.

☁️
AWS Cloud Architecture
Designing secure, cost-efficient AWS architectures with a focus on scalability, observability, and IAM best practices.
LambdaS3VPCEC2CloudWatchIAMCDK
🤖
UI & Test Automation
Building robust, scalable open-source automation frameworks used across multi-timezone teams. Expert in reducing flakiness and improving CI pipeline reliability.
PlaywrightSeleniumPythonpytestCI/CDGitHub Actions
🧠
AI & LLM Integration
Deploying and running LLMs locally and in cloud environments. Building AI-augmented workflows that reduce manual effort.
OllamaLangChainPythonAPI integrationRAG
📋
Agile Project Delivery
Leading cross-functional teams across timezones. Experienced in sprint planning, stakeholder management, and delivery cadence.
JiraValueEdgeScrumKanban
Selected work

Featured Projects

Real problems, real solutions — with outcomes that matter.

🤖
Run any LLM locally in under 10 minutes
AI Capability — Local LLM Deployment
A step-by-step framework to deploy and run open-source LLMs (Llama 3, Mistral) locally with 4-bit quantisation. Includes prompt templates, inference benchmarks, and a simple Python API wrapper.
PythonOllamaLlama 3FastAPI
View on GitHub
⚡
Reduced test flakiness · End-to-end automation
UI Automation Framework
A production-grade Python test automation framework using Playwright with Page Object Model, parallel execution, and CI/CD integration. Built for reliability at scale across multi-environment deployments.
PythonPlaywrightpytestGitHub ActionsDocker
View on GitHub
☁️
Serverless architecture · Cost-optimised AWS design
Cloud Architecture Showcase
End-to-end AWS architecture demonstration covering VPC design, Lambda event-driven pipelines, S3 lifecycle policies, and CloudWatch observability dashboards with cost-optimisation strategies.
AWS CDKLambdaS3VPCCloudWatchPython
View on GitHub
Technical notes

From the Blog

Practical guides from the things I build and break.

AI · LLM
How to run Llama 3 locally in under 10 minutes (Mac & Linux)
Step-by-step guide to deploying open-source LLMs locally with Ollama — no GPU required, 4-bit quantised, with a simple Python wrapper.
5 min readAI
 Testing · Automation
Playwright vs Selenium in 2025: which should you pick?
A practical comparison after building frameworks with both — performance, flakiness, CI/CD integration, and when each actually wins.
7 min readTesting
AWS · Cloud
Lambda cold starts: what actually helps in 2025
Tested 6 popular strategies for reducing AWS Lambda cold starts. Only 3 made a meaningful difference. Here's what the data showed.
6 min readAWS
AWS · Cost
AWS cost optimisation patterns that actually work
Coming soon — real examples of cost reductions from production workloads using S3 lifecycle, Spot instances, and Lambda power tuning.
Coming soon
AWS Certified Soluitions Architect Associate

Let's build
something great

Open to cloud automation consulting, AWS architecture reviews,
and full-time engineering roles in APAC and remote.
I typically respond within 24 hours.

guppikan83@gmail.com
LinkedIn GitHub
🕐 Sydney, Australia · AEST (UTC+10) · Usually replies <24h