I'm an AI/ML engineering student who enjoys turning ideas into working products rather than stopping at prototypes.

My current focus is on building systems around AI agents, RAG, intelligent memory, computer vision, and full-stack applications while improving the engineering practices required to make those systems reliable and maintainable.

• 🧠 Currently building Memora AI — an intelligent memory system for AI applications
• 🔎 Exploring RAG, AI agents, semantic retrieval, and memory architectures
• ⚙️ Interested in backend systems, APIs, system design, and applied AI
• 💻 Regularly practice Data Structures & Algorithms in C++
• 📍 Based in Delhi, India

An intelligent memory infrastructure project designed to give AI applications persistent and manageable long-term memory.

The system focuses on more than simply storing embeddings — it is being designed around the complete memory lifecycle including:

Store → Retrieve → Rank → Validate → Merge → Cleanup

Core areas: Semantic Retrieval · Memory Scoring · Duplicate Detection · Memory Validation · Lifecycle Management

Currently under active development.

A local-first AI coding assistant designed to understand an entire project rather than working with isolated code snippets.

Highlights

• Project-aware retrieval using RAG
• Local AI inference
• Repository-level code understanding
• Diagnostics and contextual assistance
• Safe code-fix workflow
• Privacy-focused local architecture

Tech: Python RAG Ollama Vector Search LLMs

A computer-vision powered productivity system designed to detect distractions and help users maintain focus while studying or working.

Highlights

• Real-time computer vision monitoring
• Phone and distraction detection
• YOLO-based object detection
• MediaPipe integration
• Productivity analytics
• Desktop application workflow

Tech: Python Computer Vision YOLO MediaPipe

A modern full-stack clinic management platform built around structured healthcare workflows and multi-user functionality.

Highlights

• Modern SaaS dashboard
• Patient and clinic workflows
• Database-backed application
• Full-stack architecture
• Responsive user interface

Tech: Next.js TypeScript Prisma Database

Python · C++ · JavaScript · TypeScript

Machine Learning · RAG · LLMs · Computer Vision · YOLO · MediaPipe

FastAPI · Node.js · REST APIs

React · Next.js · HTML · CSS

PostgreSQL · Git · GitHub · VS Code

AI Agents
├── Tool-using systems
├── Retrieval-Augmented Generation
└── Context & memory management

AI Memory Systems
├── Semantic retrieval
├── Memory scoring
├── Duplicate consolidation
└── Intelligent forgetting

Software Engineering
├── Backend architecture
├── API design
├── System design
└── Production-ready development

I try to build projects that answer more than:

"Does it work?"

The questions I care about are:

• Is the architecture understandable?
• Can the system handle real-world edge cases?
• Is the code maintainable?
• Can another developer run and understand it?
• Does the project solve an actual problem?

Memora AI

Building an intelligent memory lifecycle for AI applications with semantic retrieval, memory quality validation, duplicate consolidation, and automated cleanup.

The goal is to explore how AI systems can decide not only what to remember, but also what to merge, prioritize, and eventually forget.