Headline

AI/ML Engineer | NLP & Generative AI | PyTorch • Python • Transformers | Building Production ML Systems

About

Hello World, Thank you for stopping by! I build machine-learning systems that turn models into usable software. I’m an AI/ML engineer with hands-on experience across NLP, generative AI, computer vision, and backend engineering. My work spans the full path from processing data and integrating models to evaluation, inference, APIs, and deployment. My primary technical focus is Python, PyTorch, TensorFlow, Transformers, NLP, LLMs, semantic matching, and ML inference, with backend and systems experience in FastAPI, Rust, Node.js, PostgreSQL, RabbitMQ, Docker, and distributed systems. Selected Projects: 1. AutoGrade — an event-driven OCR/ML system for evaluating handwritten answers, using asynchronous processing and persistent storage. 2. ATS Resume Analyzer — an NLP/LLM workflow for resume parsing, entity extraction, semantic skill matching, and AI-generated feedback. 3. ProtonPay — a payment-system architecture combining transaction orchestration, tamper-resistant storage, and AI-assisted fraud/anomaly analysis. I’m particularly interested in NLP, generative models, ML systems, efficient inference, and the engineering required to take AI from experimentation into reliable products. I’m currently looking for opportunities in machine learning engineering, applied ML, NLP, and generative AI.

Location

Bengaluru

AI stack

claude

codex

deepseek

gemini

github copilot

kimi

Work experience

AI & ML Intern / Backend Developer at Krytil Pvt Ltd, Nov 2025 - May 2026 · 6 mo, Bengaluru, • Built AI and backend workflows across 3 products using Python, FastAPI, Node.js, TypeScript, REST APIs, and Docker, working across ML inference and application-service layers • Led the ATS Resume Analyzer: implemented resume parsing, NLP entity extraction, semantic skill matching, and Ollama-powered feedback; deployed containerized FastAPI inference for repeatable service execution. • Built a test platform for automated test creation, scheduling, grading, and progress tracking using React/Vite, Node.js/Express, MongoDB, JWT, and RBAC across 3 user roles • Built FastAPI services for AI cover-letter generation, profile management, template rendering, and ML integration for CV.KRYTIL.COM

Education

Bachelor of Engineering, Artificial Intelligence, MVJ College of Engineering, Bangalore, India

High School, High School/Secondary Certificate Programs, Bethany High School - India

Public links

github: https://github.com/joshuaroyar

joshuaroyar.vercel.app: https://joshuaroyar.vercel.app/