Ary Kashid
AI & Data Science Engineer
Engineering AI systems that see problems before they happen. Specializing in autonomous infrastructure resilience, time-series forecasting, and multimodal retrieval systems.
I build data-driven systems that turn complex problems into practical, reliable solutions.
I'm an Artificial Intelligence & Data Science undergraduate focused on machine learning, data science, and intelligent systems — working across the full path from raw data and experimentation to models, applications, and deployment. My interests span time-series forecasting, NLP, retrieval-augmented systems, and applied machine learning. I'm particularly drawn to building systems that are not only accurate, but useful, explainable, and designed for real-world constraints.
01 // ML & DATA
02 // ML METHODS
03 // ENGINEERING & INFRASTRUCTURE
Predicts infrastructure failure before it occurs using autonomous workload forecasting.
Autonomous CPU workload forecasting and infrastructure decision system using an LSTM + XGBoost ensemble, achieving an RMSE of 0.1337 and a Decision Engine v5 state machine with hysteresis and confidence-gated escalation.
Answers questions from video-based learning content via semantic retrieval and transcription.
A Retrieval-Augmented Generation system answering questions from video-based coursework, featuring an end-to-end pipeline covering Whisper ASR audio transcription, vector embeddings, and reciprocal rank fusion retrieval.
B.Tech in Artificial Intelligence & Data Science
Government College of Engineering, Kolhapur
Token of Appreciation — Sentiment Analysis Project
National-Level Hackathon · Awarded for end-to-end NLP pipeline and high accuracy ensemble model
Guard of Honor — CrashGuard AI
D.Y. Patil Ideathon · Awarded for autonomous failure forecasting architecture and real-time mitigation demo
Let's talk about ML, data & opportunities.
Open to AI & Data Science engineering roles, internships, and collaborative research initiatives.
Available for
internships, research & ML engineering roles.