Free online community for data builders. Learn data, ML, MLOps, LLMs, RAG, and AI engineering. Join free hands-on courses, events, workshops, and podcasts.
Joined October 2020
- Tweets2.6K
- Following6
- Followers11.4K
- Likes1.3K
DataTalksClub retweeted
How can people understand, evaluate, and appropriately rely on AI advice for high-stakes decisions?
Daan Di Scala researches how structured knowledge can make LLM systems more trustworthy in agrifood, healthcare, and legal.
ALT Daan Di Scala, an AI researcher, discusses GraphRAG applications in high-stakes decision-making for a webinar.
DataTalksClub retweeted
A churn model with 80% accuracy sounds good.
Here comes the uncomfortable check:
A dummy model that predicts no customer will ever churn has 73% accuracy, because only 27% of the customers actually churned.
The improvement over doing nothing is much smaller than you thought.
ALT The model predictions show an 80% accuracy rate for churn, while a dummy model predicts 73% accuracy with only 27% actual churn.
DataTalksClub retweeted
Tomorrow we are doing a live Q&A before LLM Zoomcamp 2026 starts.
If you are thinking about joining the course, this is a good time to ask questions before the cohort begins.
We will cover:
- What you will build during the course
- Who the course is for
- What prerequisites you need
- How homework, deadlines, leaderboard, peer review, and certificates work
- What to expect from the capstone project
- How to use the @DataTalksClub community during the cohort
LLM Zoomcamp is a free, hands-on course on building practical LLM applications with RAG, agents, vector search, evaluation, monitoring, and more.
The 2026 live cohort starts on June 8.
The Q&A is tomorrow, Tuesday, June 2, 17:00 CET, on YouTube.
Register here: luma.com/z25lskik
Bring your questions.
ALT Details for the LLM Zoomcamp 2026 Pre-Course Q&A featuring course information, prerequisites, and topics for discussion.
DataTalksClub retweeted
5 main AI agent evaluation dimensions:
1. Goal -> Action
2. Goal -> Plan
3. Plan -> Action
4. End-to-end consistency
5. Efficiency
🧵
DataTalksClub retweeted
Milla Jovovich (@MillaJovovich) has unexpectedly entered the AI memory conversation with MemPalace, an open-source project she launched with developer Ben Sigman (@bensig).
MemPalace argues that long-term memory for AI assistants does not have to mean more model calls, more infrastructure, and more complexity. It can be local, deterministic, and cheap to run.
The project is inspired by the classical "memory palace" technique: instead of treating memory as a flat retrieval problem, it organizes information into spatial structures.
I wrote a short explainer about why this project stands out and what it suggests about a different path for AI assistant memory:
DataTalksClub retweeted
Most discussion around AI coding agents focuses on what they can do.
Less on how to evaluate them properly.
We're hosting a workshop with @ErnstHaagsman from @jetbrains on evaluating AI coding agents using @teamcity and SWE-bench.
👉🏼 luma.com/f6zc5lr6
ALT Webinar announcement on evaluating AI coding agents with Ernst Haagsman, featuring registration details for April 28 event on YouTube.
DataTalksClub retweeted
Here's why I recommend learning in public:
Dimitris Zacharenakis has been an active participant in the Data Engineering Zoomcamp.
He found several issues in the course material, pinged me about them, and we fixed key problems thanks to his feedback.
His contributions made the course better for everyone.
At the same time, he's been posting about what he's learning.
The results: 56,200 impressions, +168% vs last year, 16,000+ people reached.
When you learn in public, you get two things:
1. You help improve the materials and community
2. You build your own visibility and network
Thanks to Dimitris and all other participants who share their progress and report issues.
If you're learning something right now, start sharing it.
ALT A graph shows a significant increase in content impressions, highlighting the impact of sharing learning experiences publicly.
DataTalksClub retweeted
Got this message from a Data Engineering Zoomcamp participant.
John-Luke landed their first junior DE role while still working on the final project.
Always great to see participants finding real value in the material!
ALT A user named John-Luke expresses gratitude for a course that helped him secure a junior DE role, while completing his project for certification.
DataTalksClub retweeted
How to evaluate AI agents step-by-step?
1. Verify goal understanding
2. Assess plan quality
3. Inspect tool execution
4. Compare plan and execution
5. Evaluate replanning
6. Measure efficiency
7. Review end-to-end consistency
🧵
DataTalksClub retweeted
In 4 days, I will be live on the @DataTalksClub podcast, taking live QAs on AI Engineering, Skill Stack, Agents, LLMOps, and How to Ship AI Products.
Together with @Al_Grigor we will chat about:
The principles behind production-grade systems
How to build reliable agentic workflows
The LLMOps practices that keep models stable
The patterns I've learned from building more than 20 AI applications
Plus answering all your amazing questions!
The podcast is live, virtual and FREE on 26 Jan.
Join here: luma.com/fx3wplwn
See you there ✌️
Thank you @Al_Grigor for inviting me! So excited about this one.
DataTalksClub retweeted
Join my 7-day free email course on AI agents!
Learn to build a full AI agent from any GitHub project, covering:
🔸 Step-by-step email guidance
🔸 Hands-on exercises
🔸 Completion certificate to flaunt your skills!
Ready to dive in?
1/3
DataTalksClub retweeted
Why the #MLOpsZoomcamp?
Several courses claim to teach MLOps, but @DataTalksClub Zoomcamp stood out for these reasons:
1. Project-Driven Learning
2. Community and Accountability
3. Completely free course
4. Well curated and in-depth course having a prevalent curriculum.
DataTalksClub retweeted
I just received "Fundamentals of Data Observability" by @noootsab.
Andy introduces the best practices of data observability and explains how to implement them in your project.
Read the book for free here: kensu.io/oreilly-all-chapter…
Thanks, Andy, for sending me a copy!
DataTalksClub retweeted
Already 4,300 people signed up for LLM Zoomcamp!
You can join us too.
Register to be the first who's notified about the course start.
Join the course: airtable.com/appPPxkgYLH06Mv…
DataTalksClub retweeted
Great news! Over 50,000 people joined @DataTalksClub from all over the world.
A big thank you to everyone who's part of our community.
Join us by registering at datatalks.club/slack.html
Thanks again for being with us. We're excited to keep growing with your help!
DataTalksClub retweeted
A fascinating update about the DTC Zoomcamp Q&A Challenge prizes!
Nvidia is giving us a GPU (RTX 4080) as the prize for the competition.
Participate today, only 20 days are left!
kaggle.com/competitions/dtc-…
DataTalksClub retweeted
We have a new book of the week – "Distributed Machine Learning Patterns" by @TerryTangYuan.
You can talk directly to Yuan, the author of the book, and win a free copy
You can join until Jan 18 👇🏼
DataTalksClub retweeted
Last week on the ML Zoomcamp, we announced a new NLP Kaggle competition.
Timur Kamaliev, our teaching assistant, prepared a getting-started notebook for the competition. You can check it at github.com/DataTalksClub/kag…
Don't forget that you can use @saturn_cloud's free GPUs.
DataTalksClub retweeted
We started the deep learning module on the Machine Learning Zoomcamp.
We recommend using Saturn Cloud for this part of the course.
We partnered with Saturn Cloud and now have a special offer for our students. 👇🏼
DataTalksClub retweeted
Join us this week on @DataTalksClub for a chance to chat with Rui Machado and Hélder Sousa Russa, authors of "Analytics Engineering with SQL and DBT" – and win a free book!
datatalks.club/books/2023110…