Hi, I'm

Dhanush Chinivar

Full-stack & AI Engineer

3+Experience
5+Projects
AI & Full StackFocus
Open to WorkStatus
Open to RelocateAvailability
scroll
Dhanush Chinivar
Dhanush Chinivar
Dhanush Chinivar

About me

Building things
that think.

I'm a Full Stack and AI engineer in my final semester at Deakin University, with over three years of industry experience building real-world software. At my core, I'm an extremely curious learner who enjoys breaking things down and building them back better.

I care deeply about writing clean, maintainable code and creating work that has real impact, from designing scalable systems to experimenting with LLMs, agentic workflows, and data-driven solutions.

Outside of work, you'll find me at the gym, out hiking around Melbourne, or exploring new cafés and cuisines. I enjoy traveling and discovering new cities, or catching something at the cinema. I'm also a big dog person. If I get distracted mid-conversation by a golden retriever, that's entirely on brand.

Career experience

Experience that shipped.

Teaching Assistant

Code CampMelbourne, AU
Apr 2025 – Present
  • Facilitated immersive multi-day JavaScript and coding camps, guiding students through 20+ hours of hands-on programming activities.
  • Mentored students in problem-solving and programming fundamentals, improving engagement and confidence.
  • Supported structured sessions to build foundational coding concepts across diverse skill levels.

Application Developer Intern

NubevestMelbourne, AU
Jan 2025 – Apr 2025
  • Conceptualised and designed custom CRM solutions, e-commerce apps, and payment platforms for small businesses, resulting in a 50% increase in user retention and a significant decrease in bounce rates.
  • Redesigned client websites, improving UI and reducing load times by 40%, which increased conversions by 70%.
  • Ensured product alignment with client expectations by conducting regular beta version review calls.

Software Engineer I

Microchip Technology Inc.Bengaluru, India
Aug 2022 – Jun 2024
  • Adopted Evolutionary Prototyping SDLC for developing software libraries for 16-bit dsPIC devices for MPLAB® Code Configurator.
  • Performed routine code reviews and mentorship, reducing production bugs by 25% and fostering a collaborative team environment.
  • Collaborated with framework engineers to implement JSON-based plugin features, enhancing user workflows and increasing adoption rates by 25%.
  • Managed end-to-end development of an ADC module, delivering on schedule with 98% code coverage and zero post-deployment issues.
  • Developed automated batch scripts that reduced manual processing time by 85%, freeing up 15+ developer hours weekly.

Software Engineer Intern

Microchip Technology Inc.Bengaluru, India
Jan 2022 – Jul 2022
  • Executed high-priority bug fixes and feature enhancements, contributing to on-time delivery of a major device release.
  • Leveraged algorithm development, debugging, technical analysis documentation, code review, and unit testing.

Projects

Selected work, laid out clean.

01

MealsForge

"Imagine tracking your macros manually every week, writing grocery lists by hand, and still blowing your budget by Wednesday. MealsForge eliminates that chaos — an AI that designs your personalised weekly plan, auto-builds your shopping list aisle-by-aisle, and keeps your streak alive with adherence analytics and nutritional nudges, all behind a frictionless subscription."

Next.jsTypeScriptPostgreSQLPrismaClerkStripeOpenAI
02

Rallify

"Tracking stocks manually, missing price swings, and reading generic market news that has nothing to do with your portfolio. Rallify fixes that — a real-time stock platform with live candlestick dashboards, custom price-alert thresholds, and an AI Daily Digest that hits your inbox every morning with a personalised market summary built from your own watchlist."

Next.jsTypeScriptMongoDBBetter AuthFinnhubGeminiInngestResend
03

Role-Aware RAG Chatbot

"Imagine dumping your entire company knowledge base into a chatbot, then realising any employee can now ask it to summarise the CFO's salary review or pull a confidential security audit. Role-Aware RAG Chatbot kills that risk — it semantically routes every query to the right knowledge domain (finance, HR, engineering, marketing), then hard-checks the user's role permissions before a single token is returned. Right data for the right role. A clean 403 for everyone else."

FastAPINext.jsTypeScriptPythonQdrantGroqSentenceTransformersTailwindCSS
04

Multimodal RAG

"Most RAG pipelines treat images as an afterthought convert them to captions, lose half the detail, and call it multimodal. This system skips the middleman entirely. By embedding both PDF text and images into CLIP's shared vector space, a plain text question can retrieve the exact chart, diagram, or figure it's asking about — no captions, no information loss. Claude then reasons over the retrieved text and visuals together to produce answers that actually reflect what's on the page."

PythonFastAPIStreamlitCLIPQdrantClaudePyMuPDFBGE
05
In Progress

AI Business Intelligence Agent

"Small business owners drown in spreadsheets, Notion pages, and inboxes — then still can't answer 'which product made me the most money last quarter?' AI BI Agent changes that. Ask a plain English question, and a four-agent LangGraph pipeline (Data → Analysis → Insight → Action) pulls live data from Google Sheets, Notion, and Gmail via MCP, reasons across all of it, and delivers a real answer — not a dashboard you have to decode yourself. Every Monday at 8am, n8n fires an automated report before you've had your coffee."

PythonLangGraphFastAPIMCPn8nNext.jsDocker

Certifications

Credentials & courses.

awsCloudPractitionerFOUNDATIONAL

AWS Certified Cloud Practitioner

Amazon Web Services (AWS)

Issued May 2025·Expires May 2028
AWS CloudCloud ComputingCloud ServicesAWS CertificationCloud Platform
View Credential
CompTIASec+CERTIFIED

CompTIA Security+ ce Certification

CompTIA

Issued Feb 2025·Expires Feb 2028
ID: COMP001022721616
Network SecurityCryptographyThreat DetectionRisk ManagementFirewall ConfigData Security
View Credential

Writing

Thoughts & writeups.

Medium·Apr 2026·5 min read

How I Built a Multimodal RAG System That Reads Charts and Images Using CLIP

Standard RAG pipelines ignore visual content entirely. Here's how I used CLIP's shared vector space to embed both text and images — so a plain English question can retrieve the exact chart or diagram it's asking about.

AILLMAgentic RAG

Get in touch

Let's work together.

Open to full-time and internship roles in full-stack or ML/AI engineering. Reach out on any platform below.

Currently based in

📍

Melbourne, VIC

Australia · AEST (UTC+10)