🚀 Full-Stack Developer | 🤖 AI & LLM Systems Enthusiast | ☁️ Cloud & Systems Explorer

• 🎓 CSAI undergraduate with strong foundations in Data Structures, Algorithms, and System Design
• 🌱 Currently learning Full-Stack Web Development, Cloud Platforms, and LLM-based systems
• 🔭 Building projects involving AI systems, scalable web apps, and applied ML
• 💡 Interested in Retrieval-Augmented Generation (RAG), Generative AI, and distributed systems
• 🛠️ Comfortable with both problem-solving and production-ready engineering

• Built a Retrieval-Augmented Generation (RAG) chatbot grounded in Wikipedia knowledge
• Implements hybrid retrieval (dense + sparse) for accurate context fetching
• Designed to reduce hallucinations and improve factual reliability
• Explored Knowledge Graphs and Mixture-of-Experts (MoE) concepts
• Focused on scalable LLM pipelines and real-world information retrieval

🌐 Live: WikiGPT

• Built a full-stack AI system that converts Slack threads, emails, and documents into structured Business Requirements Documents (BRDs)
• Designed a multi-agent LLM pipeline to generate 7 BRD sections in parallel with validation for contradictions and missing requirements
• Implemented a two-phase classification engine (heuristics + LLM) to filter noise and extract high-signal product insights
• Integrated Slack & file ingestion pipelines with traceability from requirement → source message
• Enabled human-in-the-loop editing, versioning, and section locking for production-grade workflows
• Export support for Markdown, HTML, and DOCX for stakeholder-ready delivery
• Focused on scalable system design, real-time pipelines, and collaborative UX

🌐 Live: https://beacon-brd.vercel.app/

💻 Repo: https://github.com/DevAryanSin/BeaconBRD

• Developing an AI-driven creative assistant for:
  • Story and narrative generation
  • Character and world-building support
• Uses advanced prompt engineering, context memory, and creative constraints
• Emphasis on:
  • Long-context handling
  • Style consistency
  • Controlled generation
• Built as a modular full-stack AI system

🌐 Live: NovelAI

• End-to-end web apps using Next.js, Node.js, MongoDB
• Integrated LLMs, embeddings, vector databases, and APIs
• Designed with production readiness and scalability in mind

• 🧠 LLM Systems: RAG, embeddings, vector search, hallucination reduction
• ✍️ Generative AI: narrative AI, creative assistants, prompt engineering
• 🌐 Full-Stack Engineering: scalable, real-world applications
• ☁️ Cloud & Systems: deployment, backend architecture, and system design

• 💼 LinkedIn: https://www.linkedin.com/in/aryansingh97/
• 🌐 WikiGPT: https://www.wikigpt.app/

⭐ If you like my work, consider starring my repositories!