ML Engineer & Graduate Researcher

Casual Kalotra

Building intelligent systems at the intersection of machine learning, NLP, and real-world impact. Currently pursuing Applied Machine Intelligence @ Northeastern.

View Projects Get in Touch Resume ↗
the spice awaits

About me

Passionate about AI
that actually ships.

I'm a graduate student in Applied Machine Intelligence at Northeastern University (GPA 3.95/4.0), specializing in ML systems, recommendation engines, LLMs, and conversational AI.

My work spans the full ML lifecycle — from distributed data pipelines and model training to deployment and explainability. I'm especially drawn to entertainment tech, personalization, and real-time user experiences.

3.95
GPA at Northeastern
2
Years in AI / ML
2
Degrees — B.S. IT & M.S. AI

What I work with

Skills & Stack

🧠
Machine Learning & AI
PyTorchScikit-learnTensorFlow Random ForestLogistic RegressionSVD / Matrix FactorizationClusteringFine-tuningSHAP / XAI
💬
LLMs & NLP
Claude APIPrompt EngineeringConversational RecommendersCollaborative FilteringNLPIntent Parsing
⚡
Big Data & Engineering
Apache SparkDistributed ComputingFeature EngineeringData WranglingFastAPISQLite
📊
Visualization & Analysis
TableauPower BIComputer VisionEDASHAP Plots
💻
Programming & Stack
PythonSQLReact 18ViteFastAPITMDB APIRender
🚀
Project & Delivery
Agile / ScrumML PipelinesGPU TrainingModel DeploymentFull-Stack

What I've built

Featured Projects

🗺
01 · Boston, MA
LA Crime Prediction System
Built ML models on real-world LAPD crime data to predict high-risk locations and peak hours via spatial-temporal trend analysis. Benchmarked Logistic Regression vs Random Forest with GPU-based training pipelines.
Apache SparkTableau Logistic RegressionRandom Forest
74.39% accuracy · Team Lead
Live · Deployed
🎬
FastAPI React 18 Claude API SVD SHAP TMDB
02 · Northeastern University · May 2026
Live on Render
Lumière
Project Cinema — Explainable Movie Recommender
A full-stack personalized movie recommendation system where users rate films and receive intelligent, explainable recommendations that improve over time. An LLM layer (Claude API) handles natural language input and explains every recommendation in plain English — solving the black-box problem of existing recommenders.
  • Three-layer hybrid pipeline: Random Forest (cold-start) → SVD matrix factorization (collaborative filtering) → weighted ensemble fusion
  • Claude API as both input handler (natural language → structured intent) and output explainer (2–3 sentence personalized explanations)
  • SHAP TreeExplainer for local explainability; TMDB API for live poster integration
  • Full user system: ratings, reviews, taste fingerprint profile, watchlist — stored in SQLite
Claude APISVD / MFRandom Forest SHAPFastAPIReact 18 TMDB APISQLiteRender
74.39% LR accuracy · 100k ratings (MovieLens) · Phases 1 & 2 complete
Live Demo ↗ GitHub ↗
👁
03 · Surat, India
AI-Integrated Web App
Designed user-friendly interfaces and implemented backend functionalities integrating foundational AI and ML techniques to support intelligent, scalable user interactions.
PythonWeb Dev ML IntegrationUI/UX
Whitestone Infotech

Where I've been

Education & Experience

Sept 2024 – Present
Master of Science – Applied Machine Intelligence
Northeastern University, Boston, MA · GPA 3.95/4.0
Pre-PhD track focusing on machine learning, deep learning, data analytics, and AI-driven applications. Built Lumière (Project Cinema) — a full-stack deployed recommender system using SVD, Random Forest, Claude API, and SHAP. Research includes spatial-temporal ML and scalable ML systems.
Jan – Apr 2024
Software Engineering Intern
Whitestone Infotech · Surat, India
Designed UI and implemented backend functionalities; integrated AI/ML techniques into application features supporting intelligent, scalable user interactions in an Agile environment.
Jan 2021 – May 2024
Bachelor of Science – Information Technology
Auro University, Surat · GPA 8.45/10
Foundation in software development, data systems, and applied computing. Developed early interest in machine learning and data-driven product building.

Let's talk

Open to ML internships
& research roles.

Whether you're building a recommendation engine, running a research lab, or just want to connect — my inbox is open.

kalotracasual@gmail.com
LinkedIn ↗ GitHub ↗ Email ↗

Resume

Download or view my full resume — education, experience, projects, and skills.

⬇ Download PDF View Online ↗