Hi, my name is Yash Tripathi
I'm a passionate Software Engineer.

Developed an AI-driven assistant that enables users to query and analyze SQL databases (MySQL, PostgreSQL, and MSSQL) through natural language. Built with LangChain for LLM integration and Streamlit for a clean UI, the system automatically translates user prompts into optimized SQL queries, executes them securely in session-based mode (no credentials stored), and returns results in under 5–10 seconds depending on database size. Integrated QuickChart to generate interactive visualizations (bar, line, pie, doughnut, etc.), making insights more accessible for non-technical stakeholders. Designed with a modular architecture, the solution can easily extend to additional databases or LLMs with minimal changes.

Developed a full-fledged SaaS platform for AI-powered voice calling, designed as a more localized and affordable alternative to global products like bland.ai, with a focus on Indian businesses. The system enables users to create customizable AI agents with configurable voices, personalities, languages, and knowledge bases, ensuring context-aware real-time responses during calls. Key features include credit-based usage, batch calling through Excel uploads, campaign automation, and SIP integration for seamless adoption with existing telephony providers. A built-in dashboard provides call analytics, history, and exportable logs. In alpha testing, early adopters reported ~30% reduction in manual workload, showcasing its potential to streamline customer support and inquiry handling.

Led the development of a real-time monitoring system using IoT-enabled gloves to track worker activity and safety during lifts. The system captured ~30 data points at 60Hz (e.g., acceleration, temperature, EDA) via a USB hub connected to a local device, processed the data streams, and passed them through a custom neural network model to assess lift severity and compute safety scores. Built a websocket-driven pipeline to transmit processed metrics to a React.js dashboard for live visualization, enabling supervisors to monitor worker performance without internet dependency. Designed scalable data ingestion and storage in MySQL for historical analysis, and implemented user authentication with dashboards for performance insights. Managed a small team to coordinate backend, ML, and deployment efforts, ensuring smooth real-time operation in industrial settings.

Designed and built a distributed blog platform using a microservices architecture, with services for posting, commenting, and moderation to ensure content integrity and user engagement. Implemented an event-driven communication layer using a custom-built message broker to handle inter-service messaging and maintain system consistency. Developed a Node.js + Express backend with MongoDB for data storage, paired with a React frontend for a smooth user experience. Deployed the entire application on a Kubernetes cluster with Dockerized containers, showcasing proficiency in containerization, orchestration, and scalable cloud-ready deployments.

Developed Anarock's official website using Strapi, Next.js, and PostgreSQL. Led backend development, API integration, and Docker deployment. Designed and defined modular components in Strapi following its recommended development standards, enabling reusability and consistency across the platform. Explored and integrated various plugins into Strapi for enhanced functionality. Collaborated on frontend integration with Next.js, ensuring a seamless and scalable user experience.