Areeb Ahmed

AI agents need memory to be truly intelligent. In LangGraph, 📍 Short-term memory keeps chat context 📍 Long-term memory stores user knowledge Together, they make agents stateful & human-like. Read the full guide 👇 medium.com/p/understandin… #LangGraph #AIAgents #LangChain #AI

Build AI Agents That Listen to Humans LangGraph's interrupt() makes it seamless: - Approve or reject actions - Review & edit state - Oversee tool calls - Multi-turn conversations Full guide with code: @areebahmed575/langgraphs-interrupt-function-the-simpler-way-to-build-human-in-the-loop-agents-faef98891a92" target="_blank" rel="nofollow noopener">medium.com/@areebahmed575… #AIAgents #LangGraph #Python #HumanInTheLoop #AI

“Humans in the loop are crucial to agent design.” – @pirroh, @Replit My new guide shows how to build controllable AI agents with LangGraph: breakpoints, state edits, time travel, and more. Dive into the full blog here 👇 medium.com/p/build-ai-age…

Last week I shared my LangGraph chatbot demo with persistent memory. Just published the complete technical breakdown - from problem identification to production-ready implementation. Full guide: medium.com/p/solving-the-… #LangGraph #AI #Chatbot #LLM

Tech Used 🔹LangChain + LangGraph (state mgmt & summarization) 🔹 Streamlit (frontend) 🔹 Python + PostgreSQL (Neon) 🔹 Google Gemini 2.0 Flash (LLM) 🔹 SHA-256 password hashing Check it out: GitHub: github.com/areebahmed575/… Live Demo: …ot-memory-summarization.streamlit.app

Built a chatbot with persistent memory + auto-summarization Summarizes after 6 msgs, keeps last 2 in detail, stores state in Postgres — so it remembers context without blowing up cost/perf. Demo 👉 linkedin.com/posts/areeb-ah… #LangGraph #Langchain #LLM #AgenticAI #AIAgent #AI

Just published a deep dive into ReAct agents - AI systems that reason and use tools. Dive into the full blog here 👇 medium.com/p/building-rea… #AIAgents #AgenticAI #ReAct #ReActAgent #GenerativeAI #GenAI #LangGraph #Langchain #Automation #AutomationDevelopment #AI #traveling

@Laaakeli Thanks!

@areeb575 cool

Just built another agentic AI assistant transforming the way we plan travel! Powered by LangChain, LangGraph, FastAPI, and Next.js, it generates personalized itineraries based on your trip details. 🗺️📅 #AIAgent #AI #Travel #langgraph #langchain #nextjs #fastapi

@ysamaila_ Start with Python, get comfortable with GenAI basics (RAG, embeddings, prompting), then pick up CrewAI, and eventually explore LangGraph for advanced agent development.

@areeb575 Hi @areeb575 what study roadmap will you advise for someone planning to go into Agentic AI?

@vk2106 Not yet.This is part of my final year project, so it's currently private. I'll definitely open-source it soon

@areeb575 Not open sourced?

@facelessavatars Thanks! Yeah, I've vibe coded some of the frontend

@areeb575 Really Cool Areeb, did you vibe code any of it?

@bookingcom @Tripadvisor @Expedia @Airbnb @KAYAK @hotelsdotcom @Skyscanner @priceline @travelocity

@bookingcom @Tripadvisor @Expedia @Airbnb @KAYAK @hotelsdotcom @Skyscanner @priceline @travelocity

Just launched my new Travel Planner #AIAgent ! Built with #Langgraph, #Langchain, #FastAPI & a sleek #NextJS frontend, it helps users: Find hotel recommendations Get flight information Emails you the details Would love your feedback! #AI #TravelTech #GenAI #LLM

Popular Frameworks: - CrewAI founded by @joaomdmoura : Multi-agent collaboration - LangGraph founded by @hwchase17 : Graph-based agent workflows 🔍Want hands-on experience? Check my repo: github.com/areebahmed575/… #Langgraph #Langchain #CrewAI #Autogen #AIAgents #LLMAgents #LLMs

An AI agent = LLM (brain) + Memory + Planning + Tools ✨ Key Features: - Reasoning: LLM-powered decision making - Memory: Short-term (context) & Long-term (persistent data) - Planning: Chain of Thought & ReAct - Tools: API integrations & real-world actions

AI agents are transforming how we interact with technology! These autonomous systems go beyond chat—they understand, plan, and execute tasks. Check out my detailed LinkedIn post to learn more: 🔗linkedin.com/feed/update/ur… #AI #AIAgents #MachineLearning #GenAI #GenerativeAI #LLMs

I just published RAG vs Fine-tuning vs GraphRAG: Choosing the Right Approach for Your LLM Dive into the full blog here 👇 🔗link.medium.com/YQnkhRavYOb #AI #MachineLearning #LLMs #RAG #GraphRAG #FineTuning #DataScience #GenAI #GenerativeAI #RAGvsFinetuningvsGraphRAG

💡 Get started with open-source LLMs today: 🔗 github.com/areebahmed575/… #GenAI #GenerativeAI #MachineLearning #AI #LLM #OpenSource #Quantization #Llama3 #GooglePalm2 #Falcon #Langchain #100DaysOfCode

🔹 What is Quantization? It’s a technique that reduces model size (e.g., 30GB → 5GB) and optimizes performance by converting 32-bit weights to 8-bit. This makes running models like Llama 2 (13B & 70B) feasible on consumer hardware like Core i3/i5 CPUs with 8GB RAM or GPUs.

Open-source LLMs like Meta Llama 3, Llama 2, Google PaLM 2, and Falcon are free to use and ready for local deployment. But here's the game-changer: Quantization!