Loading
Articles by Tuhin Paul
Handle Agent Memory, Checkpointing, and Recovery in Multi-Agent Workflows Using LangGraph
Master LangGraph for resilient AI: Learn state management, checkpointing, and recovery for robust multi-agent workflows.
Structured Output in LangChain
LangChain's structured output: TypedDict, Annotated TypedDict, Pydantic, and JsonSchema for reliable, application-ready data.
Human-in-the-Loop (HITL) Approval in LangGraph
Build production-grade fraud detection with LangGraph's Human-in-the-Loop (HITL) approval. Automate obvious cases, escalate ambiguity, and maintain context.
Tracking Metrics for Individual Nodes in LangGraph
Treating a LangGraph agent as a black box hides the real sources of failure. By instrumenting each node with latency, error, domain-quality, and cost metrics and shipping them to Prometheus + LangSmith you turn an opaque orchestration into a measurable system you can actually improve.
Building Resilient Multi-Agent Systems: Handling Tool Errors and Timeouts in LangGraph
Build resilient LangGraph multi-agent systems by handling tool errors and timeouts. Learn to design for failure and create robust AI applications.
LangChain Agents vs. LangGraph Nodes in AI Credit Scoring
LangChain Agents vs. LangGraph Nodes: Understand the shift from autonomous LLM agents to structured, deterministic nodes for production AI in fintech.
Version Control and Change Tracking in Complex LangGraph Workflows
Master LangGraph version control: Git, LangSmith, and checkpointers for complex, stateful, multi-agent systems.
The Architecture of State and an End-to-End Agentic Use Case
Master LangGraph's state model for autonomous AI agents: state management, reducers, persistence, and human-in-the-loop for robust applications.
Dissecting RAG vs. Agentic Workflows in Syndicated Loan Operations
When a group of banks lends $500M to a corporation, one bank acts as the "Agent." The Agent is responsible for the administrative plumbing: collecting quarterly financials from the borrower, calculating covenant compliance, distributing funds, and managing exceptions. It is a notoriously document-heavy, highly regulated, and deeply manual process.
When to Choose LangGraph Over LangChain in 2026
Choosing between LangChain and LangGraph in 2026: Understand DAGs vs. cyclic graphs for AI app architecture.
In LangGraph When Tool-Calling Fails to Retrieve the Right Context
LangGraph tool-calling failures: Debugging context retrieval issues, fixing schema ambiguity, state structure, and prompt disconnects for reliable AI agents.
Migrating from LangGraph to LlamaIndex/MAF
This article demonstrates how to migrate a production LangGraph implementation to LlamaIndex or a Multi-Agent Framework (MAF), using a real-world customer support automation system as our use case.
High-Fidelity Multi-Step Reasoning Chains with Automated Feedback in LangGraph
LangGraph enables high-fidelity multi-step LLM reasoning with automated feedback for e-commerce logistics, ensuring reliable, auditable, and constraint-compliant decision-making.
Building Intelligent Airline Transaction Routing with LangGraph Conditional Edges
Build an intelligent airline transaction router with LangGraph's conditional edges for real-time risk assessment and dynamic agent routing.
Dynamic Transaction Routing with LangGraph
Build dynamic financial transaction routing with LangGraph. Optimize processing, reduce latency, and ensure compliance with stateful, conditional multi-agent systems.
Building Custom LangChain Toolsets and Document Loaders for High-Velocity Financial Streams
Build custom LangChain tools & async loaders for high-velocity financial streams, enabling real-time market analysis and arbitrage agents.
Context Window Limits in LangGraph with Real time Use Case
Master LangGraph context window limits with memory management strategies like trimming and summarization for real-time AI agents.
Why LangGraph Outperforms AutoGen and CrewAI in Financial Fraud Detection
LangGraph excels in financial fraud detection due to its stateful, deterministic, and auditable graph-based orchestration, outperforming AutoGen & CrewAI.
Implementing Stateful Evaluation Loops in LangGraph
Unlock self-correcting AI with LangGraph's stateful evaluation loops. Generate, critique, reflect, and refine for production-ready outputs.
Mastering Python dict and State Management in LangGraph: An End-to-End Guide
Master LangGraph state management with Python dicts and TypedDict. Build robust AI agents and workflows with end-to-end guidance and best practices.
Overcoming Context Window Saturation in Multi-Hundred-Page FMCG Documents
Overcome LLM context window saturation for FMCG documents with RAG, FastAPI, React, and Gemini 1.5 Flash.
Enterprise-Grade Path Parsing in Python: Securely Extracting Filename, Extension, and Directory in Cloud Systems
Securely parse file paths in Python for enterprise cloud systems. Extract filename, extension, and directory using pathlib. Prevent path traversal and data leaks.
Beyond Chatbots: Auto-Regressive LLMs as Real-Time Enterprise Forecasting Engines
Unlock real-time enterprise forecasting with auto-regressive LLMs. Predict supply chain disruptions & prevent stockouts using token-level predictions. #LLMs
Telecom Tower Assistant: Multi-Modal LLM's for Field Service Repairs
Build a multi-modal AI assistant for telecom field service using vision, text, and inventory data. Automate repairs, reduce truck rolls, and integrate with SAP.
Observability in Azure Functions: Monitoring, Metrics
Unlock the power of observability in Azure Functions! This article demonstrates how to implement end-to-end monitoring for serverless applications using Application Insights and custom metrics. Learn how a real-world fraud detection system in digital banking leveraged observability to reduce fraud losses and improve performance. Includes code samples, Bicep deployment, and Kusto queries for building dashboards and alerts.
