AI engineer. I build agentic systems that close loops — RAG, multi-agent orchestration, eval pipelines, voice AI. Everyt

github

Dimple Chouhan  • Peerlist

Angel Dial is an experimental interactive web artwork — a soft ethereal internet instrument built with Next.js, TypeScript, React, and motion-driven UI systems. The experience centers around a celestial clock-machine where users rotate the dial, open dreamlike rooms, interact with objects, leave relics, and explore a pastel, otherworldly interface inspired by creative coding, nostalgic web culture, and interactive browser art. Unlike a traditional app, Angel Dial is designed as a living digital artifact: part game, part visual instrument, part surreal internet world. I’m currently developing its smooth clock-based navigation, room transitions, sound interactions, object-inside-worlds, and soft dreamy visual system.

Angel-dial

Built an AI-assisted infrastructure monitoring and observability platform designed to simulate production-style telemetry analysis, anomaly detection, and automated incident workflows. Developed backend services using Python, Flask, REST APIs, PostgreSQL/MongoDB, and Docker to process and analyze 100+ system metrics per minute across monitoring pipelines, diagnostics flows, and health-check systems. Designed real-time monitoring workflows with automated alerting, threshold-based anomaly detection, logging pipelines, and observability dashboards to improve incident visibility and reduce simulated detection latency by ~40%. Engineered modular APIs and scalable backend architecture supporting telemetry ingestion, diagnostics tracking, system health analysis, and automation-ready remediation workflows. Built frontend dashboard systems for infrastructure visualization, monitoring insights, and operational analytics with emphasis on reliability, debugging workflows, and production-style system behavior. Focused heavily on backend engineering, monitoring systems, observability, automation pipelines, CI/CD practices, and deployment consistency using Render, Docker, and Git-based workflows.

AutoOps AI — Self-Healing Infrastructure Moni

Built and deployed an AI-powered music intelligence platform exploring the intersection of recommendation systems, personality, and music discovery. Developed a full-stack architecture using Python, Flask, React, MongoDB, REST APIs, Docker, and Spotify/Last.fm integrations to generate personalized music insights, embeddings-based recommendations, and interactive listening visualizations. Designed and integrated Auralith, an AI-driven music agent capable of analyzing listening behavior, explaining musical patterns, generating playlists, and creating personalized recommendation flows. Implemented ML pipelines for feature engineering, clustering, similarity modeling, and behavioral analysis using audio features, metadata, and listening signals across 10K+ music records. Engineered scalable backend APIs, authentication systems, modular frontend architecture, and real-time interactive visual systems including Galaxy maps, Soul Orbs, and music identity analysis workflows. Focused heavily on system architecture, recommendation quality, AI integration, observability, and production-style deployment workflows.