🌐 MCP Server for NetBox
A Model Context Protocol (MCP) server built for NetBox to enable natural language access to datacenter information via LLMs. It powers interactions through Claude, allowing users to query clusters, devices, interfaces, and network configs in plain English—bridging traditional network models with intelligent assistants.
- LLM Integration: Enables Claude to understand and fetch NetBox data conversationally
- Context-aware queries: Supports natural language access to clusters, devices, IPs, interfaces, and config data
- Built on MCP: Implements the Model Context Protocol to structure and serve model data for language models
- Enhanced observability: Designed to scale with growing infrastructure visibility needs
🎙️ The India Brief
An automated daily podcast that curates and narrates top Indian news stories across Politics, Business, Tech, and Sports. The system intelligently scrapes news sources, ranks articles by importance, and uses NotebookLM to transform them into engaging conversational content for Spotify.
- Automated web scraping engine for multiple Indian news outlets
- Smart article categorization and scoring system to identify top stories
- NotebookLM integration for converting news into natural conversational format
- Automated podcast generation and Spotify publishing pipeline
A transparent AI assistant that not only performs tasks but also reveals its thought process in real-time. Sanchaari provides insights into how the LLM reasons through each request while handling various tasks from calculations to translation, making AI decision-making more understandable and trustworthy.
- Visible thought process: Shows step-by-step LLM reasoning for each task
- Multi-functional capabilities: Handles calculations, Wikipedia searches, worldwide time checks, stock prices (via Alpha Vantage), word definitions, news updates, and translations
- Real-time API integrations: Seamlessly connects with Wikipedia, financial data, and news services
- Transparent architecture: Demonstrates how LLMs process and reason about different types of queries
🚗 Safar - Ride Sharing Platform
A comprehensive ride-sharing application that connects drivers with passengers. Features include in-app chat communication, and an intelligent matching algorithm for optimal rider-driver pairing.
- Real-time location tracking and mapping
- In-app messaging system
- Intelligent ride matching algorithm
- Rating and review system
A real-time portfolio analysis tool that helps investors calculate future average holding prices and potential profit/loss scenarios. Users can simulate different buying strategies by adjusting share quantities and price variations to make informed investment decisions.
- Real-time stock data integration using Yahoo Finance API
- Dynamic average price calculator based on existing holdings and potential purchases
- Interactive profit/loss scenario simulator with percentage-based calculations
- Deployed on Vercel for reliable and fast performance
A detailed analytics dashboard visualizing the 2024 Andhra Pradesh Assembly Elections. Built with Grafana and PostgreSQL, it provides comprehensive insights into voting patterns, candidate profiles, and constituency-wise analysis across all districts of Andhra Pradesh.
- Voting analysis: Compare EVM vs postal ballot distributions across all 175 assembly constituencies, with party-wise and region-wise breakdowns
- Constituency insights: Visualization of closest contests with top 10 margin analysis, seat distribution across different regions of AP, and party performance metrics
- Candidate profiling: Detailed analysis of AP's candidates including criminal records, assets declaration (crorepati candidates), educational qualifications, and gender representation by party
- Interactive Grafana dashboards with dynamic filters for district-wise and constituency-wise analysis
An intelligent chatbot powered by Claude 3.5 Sonnet and Retrieval-Augmented Generation (RAG) that provides accurate, context-aware answers about Nutanix technology. Hosted on Hugging Face Spaces, it combines LLM capabilities with precise information retrieval to deliver reliable technical responses.
- RAG implementation: Efficiently retrieves and synthesizes information from the Nutanix knowledge base
- Claude 3.5 Sonnet integration: Leverages advanced LLM capabilities for natural and accurate responses
- Vector database for semantic search and efficient information retrieval
- Deployed on Hugging Face Spaces for easy accessibility and sharing