Rami Krispin’s Data Science Channel

Web Scraping with n8n, Docker, and Postgres 👇🏼

The following short video introduces how to set up an n8n workflow to scrape websites.

https://www.youtube.com/watch?v=avFcPdLGbv4

Claude Code for Everyone

The following tutorial by Prof. Steven Ge provides a step-by-step guide for setting up and working with Claude Code. This includes different routes by OS (Mac and Windows) and programming languages (R and Python).

https://gexijin.github.io/vibe/

There are many things in life you wish you didn’t have to learn—but end up being grateful you did.
Bash is one of those things.

It’s not flashy like Python or R, and because it feels “simple,” many of us (myself included) never prioritized learning it properly.

Looking back, if there’s one thing I’d change in my learning journey, it would be this:
learn the CLI and Bash early—before the fancy tools.

Why? Because they’re incredibly practical. Once you move closer to data science operations (aka MLOps), Bash becomes unavoidable—and suddenly, indispensable.

I came across this Bash tutorial today that looks like a solid, beginner-friendly, in-depth introduction and looks like a good resource to start with:


📽️: https://www.youtube.com/watch?v=Sx9zG7wa4FA

Introduction to Hugging Face Inference Providers 👇🏼

Inference providers is an Hugging Face API that enables access to any open source (or open weights) LLM models on the platform. The following tutorial provides a step-by-step guide for using the API.

https://www.youtube.com/watch?v=oxwsizy1Spw

Tine-tune LLMs with Tinker 👇🏼

Tinker is a training platform that provides an abstract layer for PyTorch, enabling a low cost API for fine-tuning. The following tutorial provides an introduction to this API.

https://www.youtube.com/watch?v=zY8cPov5R6M

Finally, kicking off my new newsletter about forecasting. First topic focus on modeling a change in trend with Piecewise Regression 👇🏼

https://theforecaster.substack.com/p/piecewise-regression-for-time-series

Code is available in both R and Python

Subscribe to the Forecaster newsletter to get email notification 👉https://theforecaster.substack.com/

#datascience#forecasting#rstats#python

n8n Tutorial 👇🏼

The following tutorial provides an introduction to n8n, and it is beginner-friendly. This three-and-a-half-hour tutorial covers foundational concepts of node architecture to deploying advanced AI systems such as RAG and multi-agent systems.

https://www.youtube.com/watch?v=UIf-SlmMays

♻️ Please share if you find it useful!

#n8n#datascience#ai

My weekly newslettter is out!

I share a weekly curated update on data science and engineering topics and resources.

This week's agenda:
🔹 Open Source of the Week - the FalkorDB project
🔹 New learning resources - Deep reinforcement learning, AWS course, deep learning with TensorFlow and Keras, serverless AI agents
🔹 Book of the week - Essential Math for Data Science by Thomas Nield

♻️ Please share if you find it useful!
📌 Join 33k practitioners and subscribe to receive weekly updates.

Substack: https://ramikrispin.substack.com/p/issue-66-the-falkordb-project-essential
LinkedIn: https://www.linkedin.com/feed/update/urn:li:ugcPost:7405625379232731136/
Medium: https://medium.com/@rami.krispin/issue-66-the-falkordb-project-essential-math-for-data-science-deep-reinforcement-learning-aws-eb14b20cbe85

#datascience#ai#llm#python

Build Serverless AI Agents with Langbase

The following tutorial from freeCodeCamp provides an introduction to context-engineered agents using Langbase with TypeScript.

https://www.youtube.com/watch?v=BMt-qvrEcFY

The course will be available for free from 11 Dec 2025 to 8 Jan 2026.

https://app.datacamp.com/learn/courses/designing-forecasting-pipelines-for-production

I’m excited to announce the launch of my new course with DataCamp: Designing Forecasting Pipelines for Production 🚀

This course focuses on the operational side of forecasting and walks through the core principles of designing production-ready forecasting pipelines. 🎯

Using a real-world example — automating forecasts of U.S. electricity demand, the course covers:
🔹 Working with APIs
🔹 Designing and running experiments with Nixtla and MLflow to train, test, and evaluate forecasting models
🔹 Building automated forecasting pipelines with Airflow
🔹 Implementing logging to monitor pipeline health, track model performance, and detect model drift

While examples are implemented in Python using open-source tools like Nixtla, MLflow, and Airflow, the concepts are language-agnostic and can be applied using R, Julia, or other frameworks.

This course is ideal for practitioners who are interested in learning about forecasting operations and production.

Introduction to Databricks 👇🏼

The following tutorial by Alex the Analyst, provides an introduction to Databricks. This hands-on tutorial focuses on working with data, use SQL editor and notebooks, build dashboards, and use AI with our databases.

https://www.youtube.com/watch?v=CoqZTt528ew

Docker for Data Engineering 👇🏼

The following tutorial by Alexey Grigorev provides an introduction to Docker with data engineering applications.

https://www.youtube.com/watch?v=lP8xXebHmuE

#data#docker#dataengineering

I share open-source projects in my weekly newsletter. This week, the focus was on Durbyn, a Julia package for time series forecasting 🎯.

This project by Resul Akay is inspired by R’s forecast package and provides implementations of core time-series statistical models.

Key Features 👇🏼
✅ Durbyn supports a wide variety of forecasting algorithms, including exponential smoothing (ETS/SES/Holt/Holt-Winters), auto- and manual ARIMA, ARAR/ARARMA, BATS/TBATS
✅ Formula Interface - it provides a declarative interface for model specification with full support for tables, regressors (features in ML terminology), model comparison, and panel data.
✅ Performance - the library leverages Julia's high performance, and it is multi-threaded by default

The screenshot below is an example of setting up an ARIMA model with regressors:

Project repo: https://github.com/taf-society/Durbyn.jl

📌 Join my newsletter-> https://ramikrispin.substack.com/

#forecasting#datascience#julialan

Stanford CS224R - Deep Reinforcement Learning 🚀

This course by Prof. Chelsea Finn covers the following topics:
✅ Design, implement, and train deep RL agents using policy-gradient, actor–critic, and Q-learning approaches.
✅ Apply model-based RL techniques, including learned dynamics and planning, to improve sample efficiency and control performance.
✅ Learn from demonstrations via behavior cloning and related methods, and integrate them with RL to accelerate policy learning.
✅ Build and evaluate offline RL pipelines that learn from fixed datasets while managing distribution shift and stability.
✅ Develop agents for multiple tasks using goal-conditioned RL, meta-RL, and unsupervised skill discovery to enable adaptation and reuse.
✅ Train policies from high-dimensional inputs and continuous action spaces using modern deep learning frameworks

https://www.youtube.com/playlist?list=PLoROMvodv4rPwxE0ONYRa_itZFdaKCylL

#datascience#freecourse

My recent LinkedIn Learning course - SQL Agents with Large Language Models, hit the 5000 learners milestone this weekend! 🎉

The course covers the following topics:
✅ SQL AI agent architecture – how AI agents work and the components of a basic SQL AI agent
✅ Working with LLM APIs – how to send prompts to LLM providers like OpenAI, Google Gemini, and Anthropic Claude, and how to run models locally with Docker Model Runner using the OpenAI Python SDK
✅ Prompt engineering – how to design and optimize prompts for generating SQL, and how to automate prompt creation with templates
✅ Building the SQL AI agent – connecting all components into a complete working system

This course is beginner-friendly and assumes no prior knowledge of AI agents.


More details in the course page: https://www.linkedin.com/learning/build-with-ai-sql-agents-with-large-language-models

Great article about AI agents' memory by Paul Iusztin 👇🏼

https://www.decodingai.com/p/how-does-memory-for-ai-agents-work

#ai#llm#datascience

Deep Learning with TensorFlow & Keras 👇🏼

The following tutorial provides a hands-on introduction to deep learning with TensorFlow and Keras.

https://www.youtube.com/watch?v=1B0s3cc71a8

♻️ Please share if you find it useful!
📌 Subscribe to my newsletter: https://ramikrispin.substack.com/

AWS Full Course👇🏼

The following video provides an in-depth tutorial for AWS. This 12-hour course is a beginner-level course, and it focuses on the following topics:
✅ Core cloud concepts
✅ AWS fundamentals
✅ Working with EC2 instances
✅ VPC networking
✅ Autoscaling and load balancing
✅ Security
✅ Hands-on deployment

https://www.youtube.com/watch?v=BiBu96zE92w

Data Engineering with Databricks 👇🏼

The following tutorial by CodeBasics provides an introduction to the Databricks free edition.

https://www.youtube.com/watch?v=761SQ9Hxbic

#databricks#dataengineering#data