I'm Mayank Singh an engineering Undergrad student who likes to Experiment, Test and Fabricate.
I’m fascinated by systems that sense, think, communicate, and act. From autonomous robots and UAVs to radar, embedded systems, and intelligent machines, I enjoy learning whatever a challenge demands and building solutions from first principles.
I build things that exist in the physical world circuits, sensors, robots and then make them smart by bridging them with computer vision and machine learning. My happy place is where a soldering iron meets a Python script.
$ whoami
> mayankish
$ cat interests.txt
> Embedded Systems & Firmware
> IoT & Wireless Protocols (RF, MQTT, BLE)
> Edge AI & Computer Vision
> Sensor Fusion & Time-Series ML
> Robotics & Automation
> Model Deployment on constrained hardware
$ echo $CURRENT_STATUS
> Engineering student — building in the real world- 🔌 I prototype on Arduino, ESP32 & Raspberry Pi
- 📡 Fascinated by RF comms and wireless sensor networks
- 🤖 Building robots that can see, sense and decide
- 🧠 Deploying ML models where compute is scarce and power matters
- 🔭 Currently exploring TinyML & Edge AI
- ☕ Fuelled by curiosity and bad hostel coffee
Updated on June 6, 2026
Feel free to explore and contribute!
Real hardware. Real data. Real results.
|
Wireless sensor network collecting environmental data across nodes. Backend pipeline runs anomaly detection using time-series ML to flag outliers in real time. Highlights: OTA firmware · MQTT broker · Anomaly model · Live dashboard |
Autonomous robot using a Pi camera + OpenCV for real-time object detection and tracking. Pi handles vision logic; Arduino drives the motors. Highlights: PID tracking · Object classification · UART Pi↔Arduino bridge |
|
Custom RF mesh network built from scratch to relay sensor data across nodes — no Wi-Fi, no cloud. Designed for low-power, infrastructure-free environments. Highlights: Multi-hop routing · CRC error checking · Sub-1GHz range |
Deployed a quantised CV model on Raspberry Pi 4 for real-time inference at the edge. Containerised with Docker for clean, reproducible deployment. Highlights: INT8 quantisation · 15+ FPS on Pi 4 · REST API · MLOps pipeline |
💡 Pin your actual repos on your profile and link them here!
