role:      Final-Year Electronics & Communication Engineering Student
focus:     Embedded Systems | FPGA / RTL Design | AI & ML

embedded:  STM32 / ESP32 | Embedded C | Register-Level Firmware | Peripherals
hardware:  SystemVerilog | RTL Design | Systolic Arrays | FPGA
ai_ml:     Machine Learning | Edge AI | RAG | LLM Applications

building:  Hardware-software systems with a focus on low-level implementation
exploring: Embedded Systems | Digital Hardware | Efficient AI Inference

goal:      Build reliable, efficient systems across hardware and software

Embedded C · STM32 NUCLEO-L476RG · ARM Cortex-M4 · GPIO · UART · SPI · I2C

A low-level firmware platform built to demonstrate practical STM32 peripheral programming and hardware debugging.

• Developed firmware using register-level programming on the STM32L476RG.
• Implemented and tested GPIO, UART, SPI and I2C peripherals.
• Integrated an I2C OLED display for hardware-level output and peripheral validation.
• Configured peripheral clocks, GPIO alternate functions and peripheral registers directly.
• Used STM32CubeIDE SFR/register inspection to observe peripheral states during debugging.
• Followed a datasheet → reference manual → register configuration → hardware validation workflow.

STM32H747I-DISCO · STM32 AI Developer Cloud · TensorFlow · Neural Networks · Edge AI

An Edge AI project for estimating motor operating parameters using machine learning and deploying the model on an STM32 platform.

• Built a neural-network model to estimate motor torque and rotor temperature from motor operating parameters.
• Worked with a large motor dataset containing electrical, thermal and operating measurements.
• Preprocessed and normalized the input/output features for neural-network training.
• Developed and evaluated the model using TensorFlow/Keras.
• Explored STM32 AI Developer Cloud and STM32 Edge AI deployment for embedded inference.
• Focused on bringing machine-learning inference closer to the hardware for real-time motor monitoring.

Python · Neural Networks · MIT-BIH · ESP32 · AD8232 · MQTT · REST API

A real-time ECG monitoring system combining biomedical sensing, embedded hardware and machine learning.

• Processed ECG signals from the AD8232 ECG sensor for cardiac signal analysis.
• Trained a neural-network-based model using the MIT-BIH Arrhythmia Dataset.
• Developed a pipeline for ECG signal processing and arrhythmia classification.
• Integrated the embedded system using ESP32 hardware.
• Used MQTT and REST APIs for communication between the embedded device and software components.
• Designed the system around real-time acquisition, analysis and transmission of physiological data.

Next.js · Python · RAG · Qdrant · Groq · LLaMA 3.3 70B

An AI-powered document analysis platform that retrieves relevant tender information and generates evidence-backed vendor rankings.

• Built a RAG-based semantic retrieval pipeline for vendor and tender documents.
• Used Qdrant for vector storage and semantic document retrieval.
• Integrated Groq LLaMA 3.3-70B for intelligent evaluation and ranking.
• Generated criterion-level scores for vendors based on retrieved evidence.
• Designed the system to provide evidence-backed explanations rather than relying only on generated rankings.
• Developed the user-facing application using Next.js.

Thanks for stopping by — happy to connect! 🌿⚡