# Arhaansh Jhingan

> Quant ML Developer. BTech CSE @ VIT Vellore, BS Data Science & Applications @ IIT Madras, Vellore, India.

Arhaansh Jhingan is a machine learning engineer and researcher (computer vision, time series, NLP, quant ML). This site is a terminal UI that runs in the browser. Ask its built-in AI about projects, hackathon wins, and stack.

## Experience
- Quant ML Developer @ Mintzy (Aug 2026 - present)
- ML Research Intern @ Power Projects (May 2026 - Sep 2026)
- ML Engineer Intern @ L&T (Jun 2026 - Aug 2026)
- AIML Team @ CodeChef-VIT (Jan 2026 - present)
- AIML Team @ IEEE-CS VIT (Jan 2026 - present)

## Projects
- [Fatigue Detection](https://github.com/XArhaanshX/Multimodal-Fatigue-Detection-System): Godot driving-sim telemetry plus webcam face signals (OpenCV, MediaPipe) feed a LightGBM fatigue model that escalates haptic alerts. (Winner, Panasonic Equinox '26)
- [CTF-Dortress](https://github.com/XArhaanshX/SCOPE-Reverse-Engineering): Dortress CTF 2026, reverse-engineering track: unpacked a Themida-protected Go binary and rebuilt its source as readable Go. (Winner, Dortress CTF 2026)
- [Land Dispute Radar](https://github.com/Ujjwal160107/vivaad-radar): Vivaad Radar: links court cases to land survey numbers and scores, with evidence, whether a parcel is caught up in litigation. (Winner, Innohack 2.0)
- [SilentHeat](https://github.com/XArhaanshX/SilentHeat): Beyond 40°C: quantifying the humid-heat blind spot in India's heatwave warnings
- [Tremor Detection](https://github.com/XArhaanshX/Tremor-Detection): An ESP32 + IMU wristband that watches a Parkinson's patient's hand all day, detects rest tremor on the band itself with an FFT (no cloud), turns it i…
- [Authorship Classifier](https://github.com/Ujjwal160107/CC-Xposed): ChatPrint: tells who wrote a piece of text from its stylistic signals, with a DistilBERT classifier behind a FastAPI service. (Winner, CodeChef-VIT Internal)
- [Load Forecast Survey](https://github.com/XArhaanshX/Load-forecasting): Literature review of ML methods for electrical load forecasting: the 50 newest arXiv papers, collected and summarised by script.
- [Quant ML Challenge](https://github.com/XArhaanshX/WunderFund-Quant-Dev-Challenge): Wunder Fund Alpha Connectome: predict future price moves from 112 order-book and trade features. Code goes public after 15 Nov 2026.
- [Stream-Aware B+Tree](https://github.com/XArhaanshX/database-project): A B+-tree index in C++17 that makes ascending inserts cheaper and keeps the resulting pages full.
- [Land Delay Radar](https://github.com/Ujjwal160107/adhigrahan-radar): Adhigrahan Radar (SIH 2026, SIH26017): predicts which land-acquisition projects will miss their statutory deadlines, before they do.
- [Wafer Anomaly Audit](https://github.com/XArhaanshX/Wafer-Anomaly-Detection): Audit and repair of the Fab-7 WAFER-GUARD excursion detector (GDG wildcard, AI/ML task).
- [Code History Search](https://github.com/liualt/PRISM): PRISM (Samsung PRISM GenAI Hackathon): ask about a JavaScript repo in plain English and get matching functions across its whole git history.
- [AI Interview Panel](https://github.com/liualt/Quorum): Quorum (EcoSphere 2026 by Agora): three AI interviewers run a live engineering case while the candidate fixes a real defect in a sandbox.
- [MeghMitra](https://github.com/XArhaanshX/MeghMitra): Ankur (SIH 2026): retrieves pre-approved contingency actions from district agriculture plans to match IMD's forecast.
- [Churn Prediction](https://github.com/XArhaanshX/Churn-Prediction-Project): A churn prediction system using ML models with full data pre-processing , EDA, class balancing, and performance evaluation.
- [HFLF Resolution Study](https://github.com/XArhaanshX/HFLF-Temporal-Resolution-Study): This repository is an independent research extension of the Short-Term Load Forecasting (STLF) project.
- [Stochastic HF Load](https://github.com/XArhaanshX/Stochastic-Signal-Integration-in-HF-Load-Data): An Independent Research Framework for Evaluating Forecasting Robustness
- [SLD Segmentation](https://github.com/XArhaanshX/Symbol-Segmentation-in-SLDs): Automated Circuit Symbol Localization in Single Line Diagrams
- [Load Forecasting](https://github.com/XArhaanshX/Time-Series-Load-Forecasting): Working on predicting energy load in power grids using various machine learning architectures
- [Few-Shot Handwriting](https://github.com/XArhaanshX/Handwriting-Recognition-Few-Shot-Pipeline): A research framework for deterministically identifying specific handwritten symbols (such as lowercase 'a' and uppercase 'A') in scanned paragraphs u…
- [Deepfake Detection](https://github.com/XArhaanshX/Deepfake-Detection): Interactive deepfake detector using CNNs and GNNs to analyze visual artifacts and relational structure.
- [Partial Discharge](https://github.com/XArhaanshX/Clustering-Problem-for-Discharge-Research): How robust is partial-discharge classification (simulated, sharp point, void, surface) as the signal degrades with noise?
- [PSSE Agent](https://github.com/XArhaanshX/PSSE-Agent): Agentic AI that turns a substation spec into a validated PSS/E 36 power-system model and single-line diagram.
- [Circuit Symbol ID](https://github.com/XArhaanshX/Single-Circuit-Symbol-Identification): A research-oriented hybrid Computer Vision pipeline for one-shot localization of embedded circuit symbols inside complex engineering diagrams.

## Stack
- Python, PyTorch, scikit-learn, LightGBM, XGBoost, OpenCV, Hugging Face Transformers, Pandas, NumPy, Time-series forecasting, Computer vision, NLP, Quantitative research, FastAPI, SQL, C++, Docker, Linux

## Links
- GitHub: https://github.com/XArhaanshX
- LinkedIn: https://www.linkedin.com/in/arhaansh-jhingan-swarm/
- Email: arhaanshjhingan@gmail.com