╔══════════════════════════════════════════════════════════╗
║ ADARSH · ECE STUDENT · AIOT BUILDER ║
║ Coimbatore, India ║
╚══════════════════════════════════════════════════════════╝
B.Tech in Electronics & Computer Engineering student who got tired of building circuits that couldn't think — so I taught them to.
I sit at the intersection of embedded hardware and AI — the place where a sensor reading becomes an intelligent decision without ever touching a cloud server. My ECE background means I understand the hardware layer that most AI engineers skip. My projects mean I can ship the software layer most hardware engineers avoid.
Currently building SentinelEdge — an edge AI anomaly detector that runs TFLite inference on a Raspberry Pi and uses Gemini AI to generate human-readable fault reports in real time.
edge_ai/
├── TFLite · TinyML inference on constrained hardware
├── Edge Impulse · sensor model training
├── MPU-6050 · 3-axis vibration + IMU
└── DHT22 · temperature + humidity sensing
languages/
├── Python · edge scripts, data pipelines, ML
├── JavaScript · Vanilla JS, Vite, async DOM
└── Dart/Flutter · cross-platform mobile apps
ai_stack/
├── Gemini API · structured LLM outputs
├── Prompt engineering · fault diagnosis, data extraction
└── API integration · Supabase, REST, WebSockets
web/
├── Vite · frontend tooling
├── Supabase · PostgreSQL + Realtime
└── Floating UI · AOS · Luxon · SweetAlert2
tools/
├── VS Code · Antigravity IDE · Git
├── Edge Impulse · Raspberry Pi OS
└── Electron · desktop app packaging
|
⬡ SentinelEdge AI-powered industrial anomaly detector. Raspberry Pi → TFLite → Gemini AI → Supabase → live dashboard. Detects faults locally. Explains them intelligently.
|
📚 Smart Syllabus & Study Planner Upload a syllabus → get an AI-generated study schedule, deadline tracker, and tips. Built as both a macOS desktop app and Android mobile app — from zero prior experience.
|
# What I'm working on
focus = {
"building": "SentinelEdge v1.0 — wiring the real hardware",
"learning": "Edge Impulse model training + MQTT protocol",
"targeting": ["Bosch BCAI", "Siemens", "NVIDIA", "AIoT startups"],
"timeline": "Final year ECE → AIoT engineer",
}
# What I believe
thesis = """
The most valuable AI engineer in 2026 isn't the one
who can call an API — it's the one who understands
what the hardware is actually saying.
"""
Open to: AIoT internships · embedded AI roles · collaboration on hardware+AI projects
Location: Coimbatore, Tamil Nadu, India
[ edge node online · inference running · building in public ]