Techie dude idk
Repurposed a legacy Intel Core i5-3470 system into a high-speed network-attached storage (NAS) server running Unraid 7.1.2, integrating a dedicated Gigabit Ethernet card to maximize network throughput. Architected a fault-tolerant storage array utilizing salvaged hardware, featuring a 1TB parity drive, two 500GB storage drives, and a 128GB SSD for high-speed caching. Established secure remote access via a Tailscale virtual private network and configured SMB shares, automating continuous household media backups using PhotoSync for over 200GB of active data. Deployed a Plex Media Server to stream hosted video content locally and remotely, establishing a scalable foundation for hosting future containerized applications.
Engineered a self-powered, ESP32-based wearable patch integrating GSR, DS18B20, and DHT22 sensors to continuously track user hydration, thermal strain, stress, and fatigue. Developed a Context-Aware Neural Deconvolution Engine utilizing a Recursive Inverse Filter (RIF) to isolate true sympathetic neural drive from raw skin conductance by dynamically factoring in ambient Vapor Pressure Deficit (VPD). Programmed real-time C/C++ signal processing pipelines to calculate four physiological indices (HI, TSI, SRI, FI) anchored dynamically to a 120-second personalized resting baseline to eliminate false positives. Architected a low-latency IoT pipeline syncing telemetry to Firebase Firestore, visualized through a custom React Native (Expo Go) mobile dashboard featuring live index cards and automated alert thresholds.
Developed a contactless, real-time attendance tracking system utilizing a microcontroller paired with an MFRC522 RFID reader and an I2C LCD. The custom firmware operates over the SPI bus to authenticate, read, and parse 32-byte user name strings from MIFARE tags. Integrated the BluetoothSerial library to enable the secure, trigger-based wireless exportation of attendance logs directly to a mobile device.
Designed and simulated a multi-currency vending machine controller driven by a Mealy Finite State Machine (FSM). Modeled in Verilog HDL and validated using Intel ModelSim , the architecture processes real-time inputs for 11 snack options. It features a robust 4-state system (IDLE, ACCEPT, DISPENSE, RETURN) that seamlessly handles exact payments, refunds, and invalid currencies across five global denominations (Rupees, Euros, Dollars, Yen, and Pounds).
Designed and prototyped an energy-efficient analog circuit to accurately monitor a 12V battery’s state of charge, preventing over-discharge. Simulated in Multisim and constructed on a breadboard , the system leverages LM324 Operational Amplifiers acting as voltage comparators. It utilizes a resistor ladder and a 1N4733A Zener diode to establish a constant 5.1V reference voltage , successfully driving an array of LEDs to provide immediate visual feedback on charge thresholds up to 12V.