Skip to content

AI engineer and full-stack architect

Engineering
the signal
behind AI.

I build AI-powered systems and shape cinematic interfaces where product logic, motion, and performance move as one.

CH1
AI systems
CH2
Interface motion
B.Tech
2028

About

Turning deep engineering into memorable products

I build intelligent digital systems at the intersection of AI, machine learning, and high-performance full-stack architecture.

My work blends product intuition with systems thinking: prototype quickly, validate rigorously, and ship interfaces that feel cinematic without sacrificing reliability.

Current focus
AI-first products with immersive front-end architecture
Strength
End-to-end execution from model logic to polished UI delivery
Approach
Measurable impact, resilient architecture, and performance-first motion

Toolkit

Frontend architecture
Next.js, React, TypeScript, Tailwind CSS, shadcn/ui, App Router
Motion and immersive UI
GSAP, ScrollTrigger, Framer Motion, Lenis, Three.js, WebGL
AI and data science
Python, Machine Learning, Data Science, LLMs, RAG, Data Pipelines
Cloud and systems
AWS, Azure, Google Cloud, Docker, Kubernetes, CI/CD
Backend and workflow
Java, Django, Node.js, REST APIs, WebRTC, Arch Linux, Git

Off the keyboard, I played competitive badminton for my school team and train for strength and mobility. Both run on the same loop as engineering: repeat, get feedback, recover, go again.

Experience

Where I've shipped

  1. Jul 13th 2026 - Present

    AI Engineer Intern

    Sennovate Inc.Hybrid · Coimbatore

    Building AI-powered features and internal tooling with the engineering team.

  2. Dec 2nd 2025 - Apr 4th 2026

    AI Engineer Intern

    Zensphere AcademyChennai

    Engineered interactive voice agents using LiveKit and Azure AI Foundry, integrating LLMs to automate complex speaking assessments and real-time linguistic feedback.

    • Integrated Azure AI Pronunciation Assessment for high-accuracy phonetic analysis and model optimization.
    • Deployed scalable AI features within Next.js frameworks.
    • Ensured platform stability via Playwright E2E testing.

    LiveKit, Azure AI, LLMs, Next.js, Django, Playwright

  3. June 2nd - 25th 2025

    Intern

    ResPro LabsChennai

    Engineered a modular Employee Management System (EMS) using SQL for real-time data analysis and reporting.

    • Streamlined HR processes by developing automated features for onboarding and attendance tracking.
    • Implemented highly efficient database schemas for robust operational reporting.

    SQL, Relational Databases, Data Architecture

Selected work

Five projects, shipped and live

StockSense dashboard with stock counts, a 14-day operations chart and a ledger reconciled badge
stocksense-eosin.vercel.appCH1

Odoo Hackathon 2026 · team of four

StockSense

Inventory software that can prove every number it shows.

Receipts, deliveries, transfers and stock counts run as validated operations on an append-only PostgreSQL ledger, so every balance can be traced and checked.

Next.js, FastAPI, PostgreSQL, SQLAlchemy, Alembic, pytest

LexiGuard legal health score of 32 with the top three clauses to push back on
lexiguard-ai-self.vercel.appCH2

PromptWars Exclusive · invite-only, top 400

LexiGuard AI

Reads a contract and tells you what to push back on before you sign.

Splits a contract into clauses, explains each in plain English, compares them with market-standard terms using vector search, and drafts fairer counter-language.

React 19, Vite, Express, PostgreSQL, pgvector, Gemini

Stadium AI Co-Pilot entry screen linking the control room, ground crew and fan portals
stadium-os-chi.vercel.appCH1

PromptWars · Challenge 4, smart stadiums

Stadium AI Co-Pilot

One AI engine for the control room, the ground crew and the fans.

A FastAPI and Gemini reasoning service behind three views: a control room for crowd density and incident triage, a mobile ground-crew portal, and a fan portal for gate guidance.

Next.js, FastAPI, Gemini, Azure App Service

Analysis of asteroid 1036 Ganymed: S-type composition, mission briefing and valuation
spectravein.vercel.appCH2

24-hour hackathon · team project

Spectravein

Ranks near-Earth asteroids as mining targets from live NASA data.

Ingests live NASA/JPL small-body data, classifies each asteroid's likely composition, estimates mission economics, and shows its orbit in 3D.

Next.js, React Three Fiber, FastAPI, scikit-learn, NASA/JPL SBDB

Breathe chapter: Pull the fragments together. Breathe.
decode-somnia.vercel.appCH1

Decode & Design 2026

Decode Somnia

A scroll-driven WebGL story about letting go of a thought before sleep.

You name what is keeping you awake, then scrolling carries it through four chapters (hold, break, breathe, sleep) in a WebGL scene with sound.

Next.js, React Three Fiber, GSAP, Lenis, Zustand, Howler

Also built

Research

Published in IEEE Xplore

Oct 15th - 17th 2025

Development of an IoT-based UAV Platform for Intelligent 360° Aerial Security Surveillance

2025 International Conference on Sustainable Communication Networks and Application (ICSCN)

A low-cost quadcopter for surveillance and reconnaissance that combines aerial imaging, IoT connectivity, cloud processing and secure communication.

Authors

  • Harshavardhan K (RMK College of Engineering and Technology)
  • Dr. E. Anna Devi (Sathyabama Institute of Science and Technology)
  • Dr. P. Josephin Shermila (RMK College of Engineering and Technology)
  • Gokul M, Dr. S. Sajithra Varun, Dr. C. Reeda Lenus
Read on IEEE Xplore(opens in a new tab)DOI 10.1109/ICSCN67106.2025.11308390
›Read the abstract

Over the past few years, with the growing interest in Unmanned Aerial Vehicles (UAVs), the development in the unmanned system has been a key technology for military and security uses. The system can perform different missions as it will assist in the monitoring in unfamiliar territories, forest conservation, spying enemy territory. Application which is designed in this research has an aim to mimic condition in war zone as a spy for enemy. After following the latest trends towards UAV creative design a light and small size UAV quad copter system was designed, assembled and tested in flight. A low cost design is targeted and make it competitive with other higher costs. Future improvements could include enhanced AI-driven automation, improved stealth capabilities, and longer battery life to further enhance mission efficiency.

Measured against the baseline design
ParameterBaselineProposedChange
Detection accuracy85%92%+7%
Data transfer speed2.0 MBps3.5 MBps+75%
Response latency1.2 s0.8 s−33%
Operational range500 m750 m+50%
Battery runtime2 h3 h+50%
Tracking efficiency78%88%+10%

Contact

Send the signal

I collaborate on AI-powered products, high-performance interfaces, and full-stack systems engineered for measurable impact.

Open to internships, freelance builds, and full-time engineering opportunities.

harshavardhan3259@gmail.com

© 2026 Harshavardhan K. Built with Next.js, three.js and GSAP.