I spent the last few months pulling 5G signals apart in GNU Radio to see how they actually sync up — now I'm pointing that same "find the pattern, then prove it" instinct at machine learning. CS student at SRM University, building toward AI/ML roles.
🔭 Building ML projects — moving from signal data to model-driven systems
🌱 Deepening my NLP and data analysis skills
🎯 Looking for AI/ML internships and collaborations
💬 Happy to talk Python, ML, algorithms, or how 5G synchronization actually works
Here's roughly the shape of how I think about a problem — turn a raw, noisy signal into something a model can actually learn from:
import numpy as np
def extract_features(iq_samples: np.ndarray, window: int = 256) -> np.ndarray:
"""Turn raw I/Q samples into features a classifier can use."""
power = np.abs(iq_samples) ** 2
windows = power[: len(power) // window * window].reshape(-1, window)
return np.column_stack([
windows.mean(axis=1), # average energy
windows.std(axis=1), # variability — flags noisy segments
windows.max(axis=1), # peak detection
])
# This is the same instinct behind the GNU Radio work below:
# noisy input -> structured signal -> something you can measure and trust.Open Source Contributor — GNU Radio · Jun 2025 – Oct 2025
My first real research problem: figure out whether a radio can find and lock onto a 5G cell fast and reliably, under bad conditions.
- Studied GNU Radio signal pipelines and tested synchronization approaches against 3GPP standards
- Built and evaluated detection algorithms (PSS/SSS), then measured how accuracy held up as signal quality degraded
- Automated the analysis in Python — what used to take manual review now runs as a repeatable pipeline, cutting analysis time by ~70%
The part that stuck with me wasn't the radio engineering — it was the workflow: noisy real-world data, a hypothesis, and a way to test it rigorously. That's the same loop I want to run on ML problems now.
| Languages | Python · C/C++ · Java · TypeScript/JavaScript · SQL |
| ML & Data | NumPy · Pandas · SciPy · TensorFlow · EDA · Feature Engineering |
| Backend & Web | FastAPI · Node.js · React.js · PostgreSQL · Firebase |
| Tools | Git · Linux · GNU Radio · Jupyter |
B.S. + M.S., Computer Science — SRM University · 2024 – 2028