LAI #65 What Happens When You Combine LangGraph, DeepSeek-R1, Function Call, & Agentic RAG

Author(s): Towards AI Editorial Team

Originally published on Towards AI.

Good morning, AI enthusiasts! Ever since we launched our ‘From Beginner to Advanced LLM Developer’ course, many of you have asked for a solid Python foundation to get started. Well, it’s here!

I’m excited to introduce Python Primer for Generative AI — a course designed to help you learn Python the way an AI engineer would.

Most Python courses teach syntax. That’s not enough. You need to think, build, and solve problems like an engineer — right from day one.

In this course, you won’t just go through Python fundamentals. You’ll build projects, use LLMs as coding assistants, and develop the problem-solving mindset that AI development demands.

Here’s what you’ll get:

• Learn Python by building real AI applications — Every concept is tied to a practical, real-world use case.
• Use LLMs as your personal coding assistants — Learn how to ask the right questions and speed up your learning.
• Think like an AI engineer — Develop problem-solving skills that go beyond just writing code.

Join the Course and start coding today!

Already comfortable with Python? You can jump straight into LLM development with From Beginner to Advanced LLM Developer. Or, if you’re ready to go all in, bundle both courses and save over $125!

— Louis-François Bouchard, Towards AI Co-founder & Head of Community

Learn AI Together Community section!

Featured Community post from the Discord

Abdibrokhim shared a dataset containing brain MRI samples. It includes real observations and conclusions from hospitals. This might come in handy if you are building something in MedTech or trying out a project in healthcare. Check it out on GitHub and support a fellow community member. If you have any questions or suggestions, reach out to him in the thread!

AI poll of the week!

It seems fairly evenly distributed, with the biggest use cases in coding and research. What interests you most about agents? Tell me your thoughts on this!

Collaboration Opportunities

The Learn AI Together Discord community is flooding with collaboration opportunities. If you are excited to dive into applied AI, want a study partner, or even want to find a partner for your passion project, join the collaboration channel! Keep an eye on this section, too — we share cool opportunities every week!

1. Ayanb1827 is working on a fully open-source personal study app/time management project and is looking for individuals with experience in AI agents, LangChain, agentic reasoning, RAG, and similar technologies within a React application. If you have experience in these areas and want to share some insights or chat, contact them in the thread!

2. Lisz.a is working on identifying novel biomarkers for different disorders with the help of informatics and is looking for people to help him with his ethical AI research. If this sounds interesting, connect with him in the thread!

Meme of the week!

Meme shared by ghost_in_the_machine

TAI Curated section

Article of the week

LangGraph + DeepSeek-R1 + Function Call + Agentic RAG (Insane Results) By Gao Dalie (高達烈)

This article outlines building a multi-agent chatbot using LangGraph, DeepSeek-R1, function calling, and Agentic RAG to enhance information retrieval and response generation. It explains how Agentic RAG improves traditional retrieval-augmented generation (RAG) by incorporating autonomous decision-making, enabling the chatbot to handle complex queries efficiently. It details the integration of research and development databases, using vector embeddings for document retrieval, and creating a workflow to manage query processing, document retrieval, and response generation. It addresses challenges like DeepSeek-R1’s lack of function call support and demonstrates solutions through text-based commands. The article also demonstrates the chatbot’s ability to autonomously plan actions, improving real-time decision-making and content generation for business or personal use.

Our must-read articles

1. Exploring LoRA as a Dynamic Neural Network Layer for Efficient LLM Adaptation By Shenggang Li

This article explores a dynamic approach to Low-Rank Adaptation (LoRA) for efficiently fine-tuning large language models (LLMs). Traditional fine-tuning updates all model parameters, which is computationally expensive. LoRA addresses this by freezing the base model and adding low-rank trainable updates. The author proposes an enhanced method, Rank-1 Sum LoRA, which decomposes updates into multiple rank-1 matrices and dynamically prunes unnecessary components based on data complexity. This approach reduces memory usage and improves adaptability. It includes theoretical insights, practical implementation with GPT-2, and results demonstrating LoRA’s efficiency in domain-specific tasks like medical Q&A fine-tuning.

2. Create Your Own AI Assistant: A Practical Guide to Multimodal, Agentic Chatbots for Everyday Use By Prisca Ekhaeyemhe

This article provides a step-by-step guide to building a multimodal, agentic chatbot capable of planning vacations, fetching real-time flight data, generating city images, and providing audio responses. Using Python, the author integrates abilities like OpenAI’s GPT-4o-mini for conversational AI, DALL-E for image generation, and SerpAPI for flight data retrieval. The chatbot is designed to handle complex tasks, such as suggesting travel destinations, providing cost estimates, and generating visual and audio outputs. It also demonstrates how to set up APIs, manage ability interactions, and create a user-friendly Gradio interface, making it accessible for those with basic programming skills.

3. Comprehensive Report on Model Context Protocol (MCP) with an Introduction to Cursor Rules By Don Lim

This article provides a detailed overview of the Model Context Protocol (MCP) and Cursor Rules, highlighting their role in enhancing AI-assisted software development. MCP standardizes interactions between large language models (LLMs) and external abilities, offering a modular, secure, and scalable framework for integrating diverse resources like databases, APIs, and file systems. It emphasizes human-in-the-loop controls, robust error handling, and extensibility, making it ideal for managing large-scale software projects. Cursor Rules, on the other hand, enable developers to define project-specific coding standards, ensuring AI-generated code aligns with workflows. MCP and Cursor Rules streamline development, improve productivity, and enhance code quality.

4. Quantum AI Computing By Mirko Peters

This article explores the transformative potential of quantum computing, focusing on its foundational concepts like qubits, superposition, and entanglement. It highlights how quantum systems differ from classical computers, offering exponential computational power for applications such as cryptography, drug discovery, and climate modeling. The article also examines challenges like qubit stability, error correction, and decoherence, while showcasing advancements by companies like Google, IBM, and Microsoft. With real-world applications across industries and ethical considerations in focus, the article underscores quantum computing’s role in reshaping technology and its implications for the future.

If you are interested in publishing with Towards AI, check our guidelines and sign up. We will publish your work to our network if it meets our editorial policies and standards.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI