Kumar Vishesh

Open to opportunities in a dynamic and challenging environment.

Skilled in Reverse Engineering and IoT systems. I break down binaries, dissect protocols, and build embedded solutions that work. From debugging firmware to crafting hardware hacks, I know how to get things done.

Let’s connect—I’m ready to bring real skills to real challenges.

Built a 3-node K3s cluster in a home lab environment, running mixed CPU architectures (ARM + x86) for flexible workload deployment.

• Tuned cluster networking, storage, and resource policies to improve reliability, reduce overhead, and ensure smooth operation under mixed workloads.
• This setup was for hosting Capture The Flag (CTF) events and was the upgrade of my previous Docker-based server.

Deployed a self-hosted cloud storage, media server, and NAS using Docker containers on a Raspberry Pi 4B (8 GB) with SSD and Gigabit Ethernet.

• Configured Nextcloud, Jellyfin, Samba, Tailscale, Cloudflared, qBittorrent, and Syncthing via Portainer for remote access with zero open ports
• Optimized system performance, resource usage, and network throughput
• Learned Docker orchestration, reverse proxy, and secure tunneling for secure, efficient home lab operation

Designed a custom PCB for an e-paper display using KiCad, handling complete schematic and layout design.

• Researched and selected suitable components by analyzing datasheets and ensuring compatibility
• Learned fundamentals of PCB design: routing, footprint creation, and design rule checks for fabrication readiness

Conducted hardware reverse engineering and IoT security research on the TP-Link Tapo C200 camera.

• Reverse-engineered camera hardware: mapped UART, dumped SPI flash, and achieved root via U-Boot exploit
• Extracted and analyzed encrypted firmware on Ingenic T31 SoC; reverse-engineered uc_convert utility and documented cloud communication
• Bypassed authentication, obtained root shell, analyzed key config files, and published findings in a technical blog post

Built a 42-key ergonomic split keyboard from scratch, focusing on custom design, firmware, and hardware prototyping.

• Designed keyboard layout in Ergogen YAML, created custom plates, and used Outemu switches with rotary encoders
• Developed QMK firmware for RP2040, enabling TRRS split and encoder support; resolved matrix mapping issues
• Assembled, soldered, and tested the keyboard; documented and open-sourced the project on GitHub

Performed hardware security analysis and embedded exploitation on the TP-Link TL-WR850N router.

• Reverse-engineered router hardware: mapped UART/SPI flash, extracted firmware with CH341A, and gained root via UART
• Analyzed firmware with binwalk; found hardcoded admin credentials, MD5 hashes, and plaintext Wi-Fi storage
• Cracked hashes with hashcat, obtained admin shell, and documented vulnerabilities like open UART and weak authentication

Developed an automated waste management system using computer vision and mechanical sorting to categorize waste into recyclable, non-recyclable, and organic materials.

• Achieved 90%+ accuracy in waste classification using Google Gemini AI
• Automated sorting process reduces manual handling of waste
• Real-time waste categorization with visual feedback
• Integrated IoT capabilities for remote monitoring

Developed a comprehensive IoT monitoring system using ESP32 and MicroPython for real-time environmental sensing and remote control capabilities.

• Achieved <100ms latency for remote sensor triggering using ESP-NOW
• Automated data collection from 6 different sensors with 60-second intervals
• Implemented seamless switching between ESP-NOW and WiFi modes
• Reduced power consumption by 40% using on-demand sensor activation
• Built Firebase integration for real-time data logging and analysis

Developed a high-performance IoT RC Car system using ESP-NOW protocol for wireless control and MicroPython for firmware.

• Implemented real-time dual-joystick control system with low latency response
• Achieved precise motor control using dual DRV8833 drivers and 16-bit PWM
• Designed modular architecture for easy hardware upgrades and maintenance

Developed a versatile wireless remote control system using ESP-NOW protocol for multi-device control and real-time communication.

• Achieved sub-100ms latency for real-time control
• Implemented multi-device switching with visual feedback
• Created a scalable architecture supporting 3+ devices
• Reduced power consumption by 40% compared to WiFi-based solutions

Developed an AI-powered IoT chatbot enabling real-time speech-to-text, AI-based responses, and text-to-speech interactions.

• Optimized 24-bit microphone input for MicroPython
• Integrated SD storage for audio logs and diagnostics
• Enhanced audio processing accuracy and system reliability

Created a centralized dashboard and automation system to manage international guest communications and volunteer coordination.

• Reduced update delays by ~50% when using WhatsApp automation
• Improved task assignment efficiency by 35%
• Enabled real-time tracking and issue resolution

Built a full-stack IoT solution for real-time environmental monitoring and analytics.

• Achieved 35% increase in data accuracy with minute-wise sampling
• Implemented user auth and CSV export for reports
• Developed interactive data visualizations

Designed a secure Android app to streamline hostel issue tracking and service request workflows.

• Enabled PIN and biometric authentication
• Boosted service efficiency by 25% and resolution time by 30%
• Integrated MongoDB for data management