Shivam kundan
PhD,
Electrical & Computer Engineering
Southern Illinois University, Carbondale
Welcome to my website!
About
I am a science fiction enthusiast and hold a BS, MS, and PhD in Electrical & Computer Engineering from Southern Illinois University, Carbondale.
Currently, I am the Technology Lead at OpenGrowth Ventures for the Kratos AI-native computer project.
Explore my website to find out more.
MY INTERESTS
Hardware-Software Interaction
This is what led me down the path of computer engineering - includes research and coursework on embedded systems like MCUs & FPGAs, Operating Systems, VLSI design, and hardware-software co-design.
Power Efficient Resource-Aware Scheduling For Heterogeneous MPSOCs
The primary focus of my MS and PhD coursework. Developed methodologies to maximize performance-per-Watt for both ARM & x86 architectures. See my publications to learn more.
Agentic AI
My current job involves R&D of an AI-native edge/personal computer, codenamed Kratos.
Local AI agents built on SLMs handle the operation of Kratos and also facilitate online agentic services from third parties. Users interact through natural language (voice) and graphics.
SELECT Project VIDEOS
Sun Tracking Smart Clock Using Raspberry Pi Pico
See the exact positions of sun and earth for any given time of day and day of the year. Visualizes altitude, azimuth, distance, speed, and a whole lot more. Accurate to the minute, no external communication needed whatsoever.
Hardware:
RP2040 MCU on a Pi Pico board
Custom-designed 2-layer PCB.
Software: Written in C and C++
Sponsored by PCBWay.
Functional Tricorder
The most advanced Tricorder device from Star Trek yet implemented.
If you ever ask yourself “is the sky more red at sunset?”, “how clean is my air?”, “how bright is my TV?”, “how busy are the airwaves?”, and seek the answer in real-time, beautifully visualized 18-bit color, then this is the device for you.
Hardware:
ESP32 MCU • RaspberryPi 4
15+ sensors • 4x custom PCBs.
Software: Written in C & Python
HoloLens + IoT Smart Room
AR application using object recognition and hand gestures to interact with real world objects such as sensors, lights, household electronics, etc. Provides a holographic interface to convert any room to a ‘smart room’.
Hardware:
Microsoft HoloLens • RaspberryPi 3B
Software:
C# & Unity 3D Engine (HoloLens)
Python (Raspberry Pi)
(ISCAS '20 Conference) A Machine Learning Approach For Improving Power Efficiency on Clustered Multi Processor Systems
A video presentation for my paper submission to IEEE ISCAS 2020 conference.
Contact Me
Contact Me
Whether you are an employer looking for awesome job candidates, or a fellow electronics/DIY enthusisast looking to collaborate, this is how you can reach me: