Electronics & Communication Engineering Student @ JNU Passionate about Embedded Systems, Machine Learning, and Linux Customization.

Frameworks: PyTorch, TensorFlow, Scikit-Learn

Audio & NLP: librosa, Wav2Vec, RoBERTa, Piper TTS

Vector DBs: ChromaDB, Ollama (RAG)

Design Tools: Ansys HFSS, CST Studio, Cadence OrCAD

Embedded: Arduino IDE/CLI, ESP32/Uno R4, R307S Fingerprint Sensor

OS: Arch Linux

DevOps: Docker (CPU/GPU), Git, GitHub Actions

• Developed a deep learning framework combining transformer-based text analysis and handcrafted acoustic feature pipelines.

• Fine-tuned RoBERTa-base and extracted 300+ prosodic/spectral features using librosa.

• Designed a rectangular patch antenna (2.4 GHz S-band) using HFSS with FR4 substrate.

• Achieved 608 MHz bandwidth and -14.44 dB return loss for mobile device study.

• Interactive TUI tool built in Rust for automated subtitle retrieval using REST APIs.

• Optimized binary size by 83% (60MB to 10MB) through LTO and binary stripping.

• Built a semantic search system for ArchWiki documentation using Ollama and ChromaDB.

• Implemented a custom ingestion pipeline to clean and chunk Markdown for high-accuracy retrieval.

AWS Cloud Intern @ Edu-Versity: Architected serverless pipelines using Lambda and S3; reduced latency by 40%.

Vocational Trainee @ Banaras Locomotive Workshop (BLW): Practical training in SCADA, Telephone Exchanges, and Locomotive systems.

Co-founder & Event Manager @ Ephirium Gaming Club: Managed 5+ tournaments for 150+ attendees (Valorant, CS:GO).