Tuhin Paul's contributions

Learn how to reduce latency and token costs in LangGraph multi-agent LLM pipelines using semantic caches, Graph RAG result caching, and state checkpointing.

Learn how Maximum Marginal Relevance improves RAG retrieval by balancing relevance and diversity in a LangGraph-based enterprise market research assistant.

Build an adaptive LangGraph RAG workflow that classifies query complexity and dynamically selects top-k retrieval. Learn how Graph RAG can improve relevant context while reducing noise and latency.

Build a scalable enterprise LLM support API with LangGraph, Graph RAG, multi-agent orchestration, async FastAPI, and stateful conversation handling.

Build a scalable Graph RAG legal discovery service using LangGraph, ChromaDB, NetworkX, FastAPI, hybrid retrieval, and stateful multi-agent workflows.

Build a LangGraph-based compliance advisor using Graph RAG and LLM agents to produce cited answers, reasoning steps, and confidence-scored audit trails.

Build an enterprise AI workflow that uses LangGraph agents and Graph RAG to generate creative ideas while grounding compliance checks in policy data.

Build a LangGraph multi-agent legal contract review proof of concept using LLM prompting, Graph RAG, state management, and a simulated fine-tuned legal model.

Build a LangGraph multi-agent system that uses a variational autoencoder and Graph RAG to detect and investigate anomalous financial transactions.

Build a LangGraph multi-agent system for enterprise support-ticket classification using a scikit-learn discriminative model, Graph RAG, ChromaDB, and persistent state.

Learn how to trace and debug LLM outputs with LangGraph, Graph RAG, Neo4j, FastAPI, and a React observability dashboard for compliance AI workflows.

Build a semantic cache for enterprise LLM pipelines using LangGraph, Redis, Neo4j Graph RAG, and FastAPI to reduce latency and API costs.

Build a token-efficient enterprise RAG pipeline using LangGraph, Neo4j Graph RAG, context compression, and structured LLM output.

Learn how to apply differential privacy to LLM fine-tuning with DP-SGD and Opacus, using a LangGraph multi-agent workflow to track privacy budgets, validate leakage, and audit training data.

Build an IT operations AI agent with LangGraph, Graph RAG, Neo4j, and FastAPI. Learn to orchestrate tool calls, preserve state, and require human approval for risky actions.

Build an enterprise inventory assistant that uses LangGraph, Pydantic, Graph RAG, and FastAPI to validate and orchestrate LLM function calls.

Build a privacy-aware multi-agent LLM pipeline using LangGraph, Microsoft Presidio, and Graph RAG to detect, anonymize, retrieve, and optionally re-identify PII securely.

Design a scalable enterprise LLM API using FastAPI, LangGraph multi-agent workflows, Graph RAG with Neo4j, and Redis caching and session memory.

Build a multi-hop RAG system for enterprise compliance investigations using LangGraph, Neo4j Graph RAG, stateful AI agents, and FastAPI.

Build an enterprise LLM documentation assistant using LangGraph, Redis semantic caching, Graph RAG, and Neo4j to reduce latency and cost while keeping answers fresh.

Build a cost-aware enterprise LLM assistant with LangGraph, model routing, Graph RAG, token limits, and usage monitoring.

Build a hybrid enterprise RAG system that uses LangGraph to route queries between chunk-level semantic search and document-level Graph RAG context.

Build a secure LangGraph-based customer support agent that detects prompt injection, routes safe requests, and protects Graph RAG data retrieval with Neo4j and FastAPI.

Build a real-time RAG financial advisory assistant with LangGraph multi-agent orchestration, Graph RAG, persistent memory, ChromaDB, Redis, FastAPI, and React.

Build a streaming multi-agent RAG system with LangGraph, FastAPI, pgvector, Azure OpenAI, and React, delivering retrieval progress, citations, and token-by-token responses.