Rishi RaiEmployee Attrition Prediction System
completed · AI and Machine Learning
A completed independent machine learning project from January 2026 that estimates individual employee attrition risk and presents SHAP explanations alongside each score.
Selected work
Explore Rishi Rai's software and machine learning projects, including completed builds, active development, product prototypes, and planned AI engineering work.
DevTrack
completed · Software Engineering
A completed coding journal for recording daily progress, tracking streaks, organizing entries with tags, and presenting learning activity through a public web interface.
Portfolio CMS Platform
active · Software Engineering
An active full-stack portfolio and content management platform in this repository, with public project and article pages plus protected administration for structured content.
NOTE-MINT
prototype · Software Engineering
A browser-based note prototype that keeps temporary notes in sessionStorage and supports PDF export without requiring an account.
EduNext Online Learning Ecosystem
prototype · Software Engineering
A TypeScript and Next.js frontend and product prototype for an online learning experience. Its scope is product exploration, not production delivery.
Enterprise RAG Reliability Workbench
planned · AI and Machine Learning
A planned engineering workbench that would explore traceable retrieval, grounded answers, citation checks, abstention, and regression testing for RAG applications.
LLM Evaluation and Observability Lab
planned · AI and Machine Learning
A planned lab that would connect versioned LLM test cases with deterministic checks, rubric-based evaluation, trace metadata, and release comparisons.
Agentic Security Operations Orchestrator
planned · AI and Machine Learning
A planned security-operations workflow prototype that would place policy checks, approval states, bounded tools, and audit records around model-generated action proposals.
Multimodal Document Intelligence Pipeline
planned · AI and Machine Learning
A planned document-processing pipeline that would study text, layout, tables, and image regions while retaining page-level provenance and reviewable extraction output.