I am an Electronics and Communication Engineering (ECE) undergraduate at S.A. Engineering College (Anna University), Chennai (Expected Graduation: 2028). My engineering focus centers on Embedded Systems, Microcontroller Firmware, and Industrial IoT, with secondary focus on Edge AI and Intelligent Hardware.

Every featured project represents authentic, auditable engineering work with dedicated visual evidence.

Context-Aware Micro-Workspace Hazard & Proximity Guard
Target: Arduino Nano (ATmega328P @ 16 MHz, 5V) • PlatformIO • Embedded C/C++
Maturity Lifecycle: Stage 4: Hardware Bench Tested (Physical Breadboard Prototype)

• Physical Embedded System: Assembled and bench-tested on an 830-point solderless breadboard with the ATmega328P Nano, 0.96" SSD1306 OLED (I2C 0x3C), HC-SR04 ultrasonic transducer, optical IR flame sensor, passive piezo buzzer, and control pushbutton.
• Sensory Arbitration: Non-blocking cooperative scheduler executing bi-level distance hysteresis ($30.0\text{ cm} \to 33.0\text{ cm}$ hold), temporal optical flame candidate persistence ($250\text{ ms}$), and proportional audible urgency modulation ($900\text{ ms} \to 180\text{ ms}$).
• Fault Detection & Memory: Detects missing echoes (NO ECHO / ZONE UNKNOWN after $800\text{ ms}$) and persists hazard counter signatures to internal EEPROM across power cycles.
• Deterministic Build: Automated PlatformIO CI build matrix verifying AVR memory constraints (35.6% SRAM, 70.5% Flash).

View Repository & Documentation →

Deterministic Virtual Microcontroller Laboratory for ECE Education
Stack: Next.js 16 • TypeScript • Zustand • Vitest (169 Tests) • Supabase • Apache-2.0
Live System: embeddedlab-os.vercel.app
Maturity Lifecycle: Stage 3: Software Tested (Browser-based simulator; 169 unit & integration tests passing)

• Register Behavioral Simulator: Models STM32-style microcontroller peripheral registers, including GPIO electrical output states, PWM timer duty/period mathematics, ADC quantization transfer equations, and UART frame transmission.
• Automated Validation: 12-challenge automated student grading engine verified by 17 Vitest test suites (169 unit and integration tests passing in CI).
• Engineering Transparency: Explicitly scoped as an educational browser simulation tool, distinguishing register behavior from physical silicon execution and analog SPICE modeling.

Launch Live Simulator  |  View Repository →

Industrial Edge AI & LoRa Strata Early-Warning Telemetry System
Stack: ESP32-S3 Firmware (C++) • TinyML • Sub-GHz LoRa (IN865) • Next.js 16 • DGMS CMR 2017
Live Deployment: mineguard-sih26025.vercel.app
Maturity Lifecycle: Stage 2: Implemented in Software (ESP32-S3 firmware & software prototype; physical hardware integration incomplete)

• SIH 2026 Prototype: Software telemetry prototype engineered for Smart India Hackathon 2026 (Problem Statement ID: SIH26025) under the Ministry of Coal / Coal India Limited.
• Firmware Architecture: ESP32-S3 driver implementations for BNO085/MPU6050 and ADS1220 24-bit strain gauge ADC, non-blocking FreeRTOS tasks, and theoretical SX1262 LoRa link budget calculations.
• Edge Intelligence: On-device running MAD/EWMA anomaly scoring with adaptive reporting cadences; benchmark achieves 93.00% recall on 1,000 synthetic geomechanical test vectors.

Launch Live Platform  |  View Repository →

Multi-Modal Early-Warning Software Demonstrator for Li-ion Battery Thermal Runaway
Stack: Vite • React 19 • TypeScript • Oxlint • WCAG 2.1 AA • Apache-2.0
Live Demonstrator: quartzsentry-prototype.vercel.app
Maturity Lifecycle: Stage 2: Implemented in Software (Academic software demonstrator with synthetic telemetry)

• Academic Simulation Demonstrator: Software simulation platform modeling early-precursor thermal runaway stages in 16-cell Li-ion battery modules.
• Multi-Modal Sensory Modeling: Mathematical representations of Electrochemical Impedance Spectroscopy (EIS Nyquist drift), ultrasonic acoustic cavitation, off-gas emissions ($H_2$ / VOCs), and thermodynamic runaway boundaries.
• Accessibility & Ethics: WCAG 2.1 AA compliance (6.16:1 text contrast) with prominent disclaimers establishing that telemetry is synthetic with no physical battery pack or certified protection hardware.

Launch Live Demonstrator  |  View Repository →

Auditable Academic Evaluation, GPA/CGPA & Admissions Planning Platform
Stack: React 19 • TypeScript • Vite • Tailwind CSS v4 • Vitest (31 Tests) • jsPDF • MIT
Maturity Lifecycle: Stage 3: Software Tested (31 unit tests passing)

• Deterministic Evaluation Engine: Computes Anna University 10-point GPA/CGPA credit-weighted evaluations with official Tamil Nadu Engineering Admissions (TNEA) 200-point reduced formulas: $\text{Cutoff} = M + \frac{P}{2} + \frac{C}{2}$.
• Export & Type Safety: High-resolution client-side PDF/CSV transcript generation and national entrance exam eligibility rulesets (JEE Main, JEE Advanced, NEET UG) verified by 31 Vitest unit tests.

View Repository & Tests →

My firmware development adheres to real-time embedded safety and deterministic execution principles:

• Low-Level Hardware Interfaces: Direct register configuration, ADC sampling calibration, volatile qualifiers, bounded timing loops, and hardware watchdog timer (WDT) refresh cycles.
• Bus Protocols: Hardware I2C (address arbitration, clock stretching, bus recovery), SPI (clock phase/polarity modes), UART (frame buffering, ring buffers), and LoRa sub-GHz packet structures.
• Noise Mitigation: Temporal persistence filtering, moving average filters, and bi-level hysteresis thresholds to eliminate contact bounce and sensor chatter.

Bridging physical sensor instrumentation with embedded machine learning:

• Constrained Anomaly Scoring: Running Median Absolute Deviation (MAD), Exponentially Weighted Moving Average (EWMA), and Robust Z-score algorithms for low-memory microcontrollers.
• Adaptive Telemetry: Dynamic reporting frequency scaling (reducing radio duty cycle during quiescent conditions; escalating packet rate during rapid sensor drift).
• Geotechnical & Battery Research: Physical link budget formulation, InSAR geospatial alignment, and multi-spectral early fault precursor identification.

AWS credentials earned through proctored examinations. Public verification links are provided below; raw AWS validation identifiers are intentionally not published in this README.

Certification	Issuing Body	Issue Date	Expiration Date	Verification Proof
AWS Certified Cloud Practitioner	Amazon Web Services	August 9, 2026	August 14, 2029
AWS Certified AI Practitioner	Amazon Web Services	August 1, 2026	August 1, 2029
AWS Certified Developer – Associate	Amazon Web Services	August 14, 2026	August 14, 2029
AWS Certified Solutions Architect – Associate	Amazon Web Services	August 7, 2026	August 7, 2029