Tejas Kumar

Shift Kuala Lumpur

How to Thrive as a Professional

This talk is by a developer and for developers but is applicable to a wider audience. In it, we will explore how professionals can navigate and have success in an industry slowly being consumed by AI and-through peer-reviewed research-identify how to ensure job security and success regardless. There will also be a technical and demo-heavy section where we practically explore how we can leverage AI to improve our productivity, careers, and experiences of ourselves and our users.

Shift Kuala Lumpur

The Complete Guide to Choosing an AI Agent Framework in 2025

This is a developer-oriented, technical, and balanced review of Langflow, n8n, OpenAI's new AgentKit, LangChain (+ LangGraph), CrewAI, and AutoGPT—plus a decision matrix with 10 factors, use case recommendations, and honest pros/cons all oriented to helping you and your team choose the right tool for your specific use case.

Compare OpenAI's AgentKit and Langflow by building an agent

The recent launch of OpenAI's AgentKit has sparked excitement among developers, but how does it stack up against existing solutions like Langflow? In this in-depth comparison, we'll dive into the strengths and weaknesses of each platform, exploring their capabilities and trade-offs through a real-world example. By building the same AI agent in both platforms, we'll help you make an informed decision about which tool is right for your next project.

Build Your Own GPT-5: Smart Model Routing with Langflow

Between o3, o1, 4o, GPT-5, Claude-4.1-opus, Gemini, etc. choosing the right model name for our applications can feel like picking a health insurance plan. In this post, we fix this with Langflow by building a powerful Multi-Agent System (MAS) that implements the LLM-as-judge pattern: one agent decides "think" or "default," an if-else routes the query, and specialized agents do the work.

How to Build a Deep-Research Multi‑Agent System

This blog post dives into Multi-Agent Systems (MAS) and explains how to build a basic deep research workflow in Langflow and explains when to prefer them over single-agent systems. We also explore how Langflow particularly shines here, enabling rapid iteration and cherry picking the best models per specialized agentic task.

How to Create Secure AI Applications

💡This post also exists as a video for those who prefer watching over reading. You’ve built an impressive AI agent. It can query databases, call external APIs, and even

LLM Observability Explained (feat. Langfuse, LangSmith, and LangWatch)

Building a new application powered by Large Language Models (LLMs) is an exciting venture. With frameworks and APIs at our fingertips, creating a proof-of-concept can tak

From 'no-reply' to 'please-reply': How to build an AI-Powered Email Assistant with Langflow and SendGrid

This post will walk you through how you can build your own "please-reply" system using Langflow, the easiest way to build AI agents, and SendGrid's Inbound Parse. We'll show you how you can build one for your own organization.

How to Host Your AI Agents and MCP Servers on Langflow Anywhere

This guide will walk you through deploying Langflow to a variety of popular hosting platforms - FlightControl, Fly.io, Render, and Hetzner - transforming your local projects into globally accessible AI powerhouses.

How to Create Your First AI Agent

Learn what AI agents are and how you can create your own with Langflow.

Instantly Chat with Your PDFs Using Langflow

Here’s a quick look at how Langflow makes it easy to build an app that enables you to chat with PDFs.