AI Product Builds & Demos
Curated by Raghu Challapilla, AI Product Manager
AI Builds
Production-ready AI applications solving real-world problems
House Whisperer
Inspectors spend 1 - 3 hrs writing reports. Multimodal multiagentic RAG system analyzes property conditions using InterNACHI ASHI standards to deliver instant, professional-grade inspection reports. Bonus: Home Inspector assitant to answer all questions about standards, codes, and guidelines
ModelGov
Built for the Bank of America Innovation Challenge, this RAG-based assistant validated that domain knowledge onboarding can be done in days rather than months with scripted NLU bots.
Asimov-Vedanta Interface
AI ethics and philosophy exploration platform. Bridges Western AI safety principles with Eastern philosophical wisdom through cross-domain reasoning and dialogue systems.
Technical Deep Dives
12+ technical demos covering advanced AI engineering topics
Guardrails and Caching for AI
Production-ready AI with security guardrails and 2-3x faster responses through intelligent caching strategies.
A2A Agent-to-Agent Protocol
Implementing Google's A2A protocol enabling independent agents from different vendors to discover and collaborate.
LangSmith Studio & Modular Agents
Building modular, deployable agent platforms with instant model/tool swapping using LangGraph Studio.
Exploring MCP: Creating Tools for LLMs
Deep dive into Model Context Protocol (MCP) - the standard for LLM clients to discover and utilize external tools and data.
Open Deep Research with Smaller Models
Cost-effective AI research application using modular graph-based pipelines to produce high-quality outcomes efficiently.
Advanced RAG Retrieval Strategies
Implementing 6 retrieval strategies: BM25, multi-query, contextual compression, parent document, ensemble, and semantic chunking.
Evaluating RAG with RAGAS
Building robust evaluation frameworks for RAG and agentic applications using RAGAS with synthetic test data generation.
Synthetic Data Generation with LangSmith
Creating synthetic test data for RAG pipeline evaluation using LangSmith for comprehensive quality assurance.
Multi-Agent LLM Systems
Building sophisticated multi-agent systems using LangGraph for complex orchestration and coordination.
Building Multi-Step Agents with LangGraph
Creating intelligent agents capable of multi-step reasoning and multi-tool coordination using LangGraph.
Exploring LangSmith: LLM Observability
Using LangSmith for monitoring performance, evaluating correctness, tracking latency, and debugging AI applications.
Building RAG with LangChain & Python
Complete RAG system implementation using LangChain, LangGraph, LangSmith, LCEL, and Python.
About
Product Leader with hands-on AI engineering skills delivering applied AI systems at enterprise scale