9 GitHub Repositories that will give you superpowers as an AI Engineer: 1. Build a Large Language Model from scratch

This repository contains the code examples for developing, pretraining, and finetuning a LLM from scratch.

It is the official codebase for the book Build a Large Language Model (From Scratch).

- Github Repo: github.com/rasbt/LLMs-fro…

AI agents don't need bigger models, they need better context!

You can now make Claude Code, Gemini CLI, and other coding agents 10x more powerful by giving them the right context.

Here’s a step‑by‑step breakdown: Context engineering is about delivering the right information and tools in the right format so LLMs can perform reliably.

Most of the time when agents fail, it’s not the model but missing context or tools that prevent reliable performance.

Let’s see how to set that up.

You can now make Claude Code, Gemini CLI, and other coding agents 10x more powerful by giving them long‑term memory!

It just takes a single line of code.

Here’s a step‑by‑step breakdown (100% local): Coding agents have a major limitation: they forget everything between sessions.

Without memory, they can’t retain project context, past fixes, or key decisions.

What they need is a persistent memory layer to store and recall context.

Let’s see how to set that up.

10 GitHub Repositories you should definitely check as an AI Engineer: 1. Hands on Large Language Models

This repository contains notebook examples that cover everything from the introduction to language models to fine-tuning them.

Check this out: github.com/HandsOnLLM/Han…

5 MCP, RAG & Agents projects for AI Engineers: 1. Agentic RAG

Build a RAG pipeline with agentic capabilities that can dynamically fetch context from different sources, like a vector DB and the internet.

Check this out: github.com/patchy631/ai-e…

Microsoft released an AI powered data analysis tool!

Data Formulator is an AI-powered tool for analysts to iteratively create rich visualizations.

It's no-code & 100% open-source Github Repo: github.com/microsoft/data…

Let's build a PDF RAG Agent, step-by-step: 1. Here’s what we’re building:

The agent pulls context from the indexed PDF files stored in vector DB. If needed, it can also fall back to web search to provide more relevant answers.

Tech stack:

@milvusio for vector DB as knowledge base
@AgnoAgi for agent orchestration

5 MCP servers that will give superpowers to your AI Agents.

100% open source. 1. Fast API

Turn any FastAPI app into an MCP server!

FastAPI-MCP is a zero-configuration tool for automatically exposing FastAPI endpoints as Model Context Protocol (MCP) tools.

github.com/tadata-org/fas…

I've been coding in Python for 5 years now.

If I were to start over today, here's a roadmap: 1. CS50’s Introduction to Programming with Python:

The course is designed for students with or without prior programming experience who’d like to learn Python specifically.

Check this out: cs50.harvard.edu/python/2022/

5 MCP servers that will give superpowers to your AI Agents.

100% open source. 1. @firecrawl_dev MCP server

Firecrawl MCP lets you scrape, crawl, search, extract, and run deep research at scale.

It supports JS rendering, batch scraping, URL discovery, and full content extraction from web search.

Let's build a RAG app over Excel sheets using Docling and DeepSeek-R1 (100% local): 1. Check out the RAG app we're building!

This app allows you to upload Excel files and chat with them using the DeepSeek-R1 model running locally.

Built entirely in Python using @getreflex

Train and Evaluate all your Machine Learning Models at once!

Introducing Lazy Predict a Python Library that allows you to train, test and evaluate all your ML models at once with just a few lines of code. Model selection is key for any ML Project which includes identify the model that generalizes well and provides the best performance.

This library can save time and effort by training and evaluating all your ML models at once

Here is using LazyRegressor for Regression Problems:

Let's build a RAG app using DeepSeek-R1 (100% Local): 1. Check out the RAG app we're building!

This app allows you to upload PDF files and chat with them using the DeepSeek-R1 model running locally.

Built entirely in Python using @getreflex

Build a LLM app with RAG to chat with GitHub using Llama 3.2 running locally on your computer (100% free and without internet): 1. Install and Import necessary libraries

@getreflex for building the web app in Python
Embedchain for the RAG functionality

Let's build a RAG app using Llama 3.2 running locally on your computer (100% open source): 1. Install and Import necessary libraries

@getreflex for building the web app in Python
Embedchain for the RAG functionality

Build a Customer Data Management Dashboard App in Python (step-by-step explanation with code): 1. Before we begin, take a look at the dashboard we're building!

This interactive dashboard displays customer data and allows you to add, edit, and delete entries in a table.

It also visualizes changes in the data over time. Built entirely in Python using Reflex:

Build a multi-agent AI news assistant using Open AI swarm and Llama 3.2 running locally on your computer (100% free and open source): 1. Install and import necessary Libraries

@getreflex for building the web app in Python
@OpenAI Swarm for multi-agent orchestration
@ollama for running llama 3.2 locally

Also:

• Load environment variables
• Initialize Swarm client
• Install and run Llama 3.2 using Ollama

Generative Pretrained Transformer (GPT) implemented from scratch using Numpy in just 60 lines of code: Although the above code is super slow and doesn't include any training code, backpropagation, GPU/TPU support, batching, etc

This is a good way to get an idea of the implementation of GPT.

Checkout the complete code here:

github.com/jaymody/picoGPT

5 Open Source frameworks to build LLM apps with RAG (100% free): 1. AutoRAG

Opensource framework tool for finding the optimal RAG pipeline for “your data.”

You can evaluate various RAG modules automatically with your own evaluation data and find the best RAG pipeline for your own use-case.

github.com/Marker-Inc-Kor…

Build a Dashboard App in Python (step-by-step explanation with code): 1. Before we begin, take a look at the dashboard we're building.

This is an interactive NBA dashboard featuring player stats built using @getreflex

The Table tab allows you to display the player data, while the Stats tab visualizes relationships between player stats.

Build a Multimodal AI Agent using Gemini 2.0 that can understand videos and search the internet at the same time (100% open source): 1. Install and import necessary Libraries

@getreflex for building the web app in Python
Phidata for building the Agent
Google AI Python SDK & Gemini 2.0 Flash model