NKM
Résumé ↓
// available for 2027 new-grad + internships
Chennai, India
B.Tech CSE '27 · CGPA 9.05
Naveen Kumar
Mohanarajan
AI / ML ENGINEER
FULL-STACK DEVELOPER
I ship models and the products around them: real-time computer vision at sub-100ms, forecasting pipelines over millions of records, and the Kubernetes platforms that keep them running at 99.9% uptime.
NAVEEN
SRM Institute of Science & Technology · 2023–2027
// 01 — technical skills
The whole
stack, spun
into orbit
Drag the sphere to spin it. Lime nodes are AI/ML and data; outlined nodes are the platform and product layers I build around them.
AI / ML
PyTorch · YOLOv8 · OpenCV · TensorFlow.js · Predictive Modeling · LLMs / GenAI
Data & analytics
EDA · Stats · Data Pipelines · Power BI · Tableau · Prometheus · Grafana
Languages
Java · Python · C++ · SQL · JS / TS
Backend & APIs
FastAPI · Express.js · REST · WebSockets · Streamlit
Cloud & infra
AWS (EKS, RDS, Route53, VPC) · Kubernetes · Terraform · Docker · Linux · OCI
Frontend
React · Next.js · Vite · Tailwind CSS · Electron
SE & planning
System Design · Full SDLC (Design→Deploy→Operate) · Agile / Scrum · CI/CD · Release Planning · Operational Excellence
PyTorch
YOLOv8
OpenCV
TensorFlow.js
LLMs / GenAI
scikit-learn
pandas
Power BI
Grafana
Python
Java
C++
TypeScript
SQL
FastAPI
Express.js
WebSockets
Streamlit
AWS EKS
Kubernetes
Terraform
Docker
Linux
React
Next.js
Vite
Tailwind
Electron
GitHub Actions
Prometheus
Tableau
Git
// 02 — education
Chennai · Jamnagar
SRM Institute of Science & Technology
2023–2027
B.Tech, Computer Science & Engineering
CGPA 9.05 / 10
Coursework: DSA, OODP, DBMS, Software Engineering, ML, Cloud Computing, DevOps
K.D. Ambani Reliance Foundation School
2021–2023
Class 12 — 80.20%
Class 10 — 95.33%
Jamnagar, Gujarat · CBSE
// 03 — experience
Shipped inside
India's biggest ops
MAY 2026 — AUG 2026
Remote
Jio Platforms Ltd.
Data Science Intern
- ▚Architected a real-time CV pipeline using YOLOv8 + OpenCV for people/occupancy analytics on live camera streams, exposing results via FastAPI with sub-100ms latency.
- ▚Developed forecasting models in Python (pandas, scikit-learn) for demand & usage, achieving 70% accuracy on held-out data to inform capacity planning.
- ▚Engineered data pipelines over 2M+ subscriber records to surface engagement trends and anomalies, directly shaping retention strategy.
- ▚Synthesized model outputs and pipeline insights into interactive dashboards and executive presentations, driving data-backed decisions across product and operations teams.
DEC 2024 — JAN 2025
On-site · Jamnagar
Reliance Industries Ltd.
ML Intern
- ▚Planned and built predictive models with 85%+ accuracy to forecast equipment lifespan, enabling a strategic shift from reactive to preventive maintenance.
- ▚Engineered end-to-end data pipelines over 5+ years of industrial data — ingestion, cleaning, feature engineering — to surface operational trends and efficiency gaps.
- ▚Delivered 3 interactive Power BI dashboards translating plant data into decision-ready insights for leadership, supporting quarterly resource planning.
// 04 — projects
Four things
that actually run
Every card carries a live preview frame — drop a screenshot into it, or hand me the URLs and I'll embed the real site.
01
FitTrack AI
Next.js
TensorFlow.js
TypeScript
Drizzle
- ▚Built an AI-powered fitness tracking platform with real-time pose detection achieving 90%+ exercise recognition accuracy using TensorFlow.js for tracking reps, form, and workout metrics directly in the browser.
- ▚Designed a gamification engine with XP-based progression, leveling, and leaderboards to drive user engagement and retention.
- ▚Architected a full-stack Next.js 14 app with Clerk authentication, Neon (Postgres) database, Drizzle ORM, and shadcn/ui component library deployed on Vercel with <200ms avg. page load.
- ▚Developed AI-guided tutorial videos and an interactive coach interface to demonstrate proper exercise form, reducing injury risk for users.
TOP 20 · BAJAJ
BEST UI/UX
02
UIUXMockUp
Next.js
WebSockets
TypeScript
PostgreSQL
- ▚Architected a real-time collaborative UI/UX design platform enabling multiple concurrent users to edit a shared canvas via WebSockets — reducing design iteration time by 40%.
- ▚Built a component library of 20+ reusable UI components against a scalable, documented design system using shadcn/ui + Tailwind CSS.
- ▚Implemented Drizzle ORM with PostgreSQL for type-safe database operations, Clerk for authentication, and React Context API for application-wide state management.
- ▚Ensured cross-device compatibility with responsive breakpoints and consistent behaviour across desktop, tablet, and mobile viewports.
03
Damaged Package Detection
YOLOv8
Python
Docker
Streamlit
- ▚Designed and trained a YOLOv8m detector for automated parcel-damage QA, identifying hole and wet defects in real time.
- ▚Architected a rules-based PASS/REVIEW/REJECT engine driven by detection count, class severity, and damaged frame area — replacing manual inspection and improving throughput planning.
- ▚Added batch processing producing annotated images and downloadable CSV/ZIP audit reports for full shipments.
- ▚Shipped a multi-page Streamlit dashboard with confusion-matrix metrics and threshold sweeps; containerized with Docker for reproducible deployment.
04
Env0
Kubernetes
AWS EKS
Terraform
Electron
- ▚Provisioned production-like dev environments in <2 min with one click — K8s namespaces + PostgreSQL, Redis, VS Code in browser.
- ▚Architected full stack: Electron desktop app + Express.js backend + K8s provisioning engine with auto-cleanup after 24h.
- ▚Deployed on AWS EKS with Terraform (VPC, RDS, Route53); CI/CD via GitHub Actions (ECR + EKS).
- ▚Achieved 99.9% uptime and <120s avg. provision time with Prometheus/Grafana + auto-SSL via cert-manager.
// 05 — achievements & certifications
Receipts
WINNER · APR 2026
Best UI/UX — CodeSprout Hackathon
Built UIUXMockUp, a real-time collaborative design platform, in a 36-hour build window.
TOP 20 · FEB 2025
Bajaj Finserv “Flex It Out”
Built FitTrackAI, a browser-based pose-detection fitness coach, from a national entry field.
// 06 — contact
Let's build
something
that thinks.
Email
naveenkumarmohanarajan38@gmail.com
Phone
+91 91066 63529
LinkedIn
/in/naveenkumarmohanarajan
PDF
Download résumé ↓
Naveen Kumar Mohanarajan · Chennai, India
AI/ML · Full-stack · 2026