The world's leading publication for data science and artificial intelligence professionals. Submit an Article ✍️ contributor.insightmediagrou…
We're Global 🌏
- Tweets 64,671
- Following 1,913
- Followers 251,546
- Likes 7,978
11 sessions, two days, one question: what does it take to trust AI agents in production? Don't miss the Agentic AI Conference 2026: free and virtual, Nov 9-10.
Keynote: “When Agents Misbehave” with Mat Velloso, Chief Product Officer at Docker.
Plus: 6 panels and 4 hands-on tutorials.
Register now to save your spot: datasciencedojo.com/agentic-…
557
Meet the speakers of the Agentic AI Conference!
26 leaders building and deploying AI agents today will be taking the virtual stage Nov 9–10 for keynotes, panels, and hands-on tutorials. They’ll take on the questions that decide whether AI agents make it out of pilot.
Check out the full speaker lineup:
datasciencedojo.com/agentic-…
1
1
2,015
The Agentic AI Conference by Data Science Dojo returns on November 9–10, 2026, with two days on the security, compliance, governance, and alignment challenges of deploying autonomous AI. The conference has drawn 178,000+ attendees across its three previous editions. Every session is recorded, so you can watch on demand if you can’t join live.
Register free: hubs.ly/Q04B0JdR0
1
5
1,538
A small open-source LLM can do more than generate text. Anubhab Banerjee breaks down how to transform Qwen into a lightweight JEV classifier designed for fast, single-pass decisions.
towardsdatascience.com/how-t…
7
1,865
As GraphRAG systems grow, not every decision needs to go through an LLM. Partha Sarkar looks at how TypeSafe Jev can handle frequent graph decisions while LLMs focus on the tasks that require deeper reasoning.
towardsdatascience.com/graph…
3
8
1,631
AI’s impact on data science goes beyond productivity gains. Yu Dong explores how more capable models are expanding the role and reshaping the career path for data scientists.
towardsdatascience.com/ai-ma…
5
1,693
Your AI bill may be growing for reasons that have little to do with simply having more users. Gursimar Singh examines how tool calls, file access, and multi-step agent workflows can turn a single request into a much larger cost.
towardsdatascience.com/your-…
2
7
1,885
What happens when you replace 1,000 small Apache Iceberg files with six larger ones? This experiment measures the impact on query performance across three different SQL workloads.
towardsdatascience.com/i-com…
1
4
1,908
What if you could turn a small open LLM into a fast text classifier? Anubhab Banerjee shows how to build a JEV model from an open-source Qwen LLM by replacing its language-modeling head.
towardsdatascience.com/how-t…
2
4
11
2,520
GraphRAG can help LLMs reason across relationships that simple vector search may miss. Partha Sarkar explores how calibrated decision models can handle high-frequency graph decisions while keeping LLMs focused on reasoning and synthesis.
towardsdatascience.com/graph…
1
2
18
2,457
AI is no longer just helping data scientists complete the same tasks faster. Yu Dong examines how it is changing ownership, judgment, and the role of data scientists itself.
towardsdatascience.com/ai-ma…
1
2
7
2,424
AI budgets are changing as companies move from simple chatbot interactions to agentic workflows that consume tokens at every step. Gursimar Singh explores why AI costs can escalate quickly and where teams may be paying twice for the same work.
towardsdatascience.com/your-…
2
3
3
2,379
Last call! Our Annual Reader Survey closes soon. This is your final chance to weigh in on the future of Towards Data Science. Thank you to everyone who's already responded — and if you haven't, there's still time: bit.ly/4bXFBGH
1
1
2,894
Our Annual Reader Survey closes soon. If it's been sitting in your tabs, now's the time. In just a couple of minutes, you can have a real say in what TDS covers, who we feature, and how we grow. Share your input: bit.ly/4bXFBGH
2
3
3,502
Thousands of data scientists, engineers, and ML practitioners read TDS every week, and we want to make sure we're serving you well. If you haven't yet, take a couple minutes for our Annual Reader Survey. Every response helps: bit.ly/4bXFBGH
5
3,036
Ever wonder how we decide what to publish? A lot of it starts with you. Last year's reader survey helped us sharpen our coverage and bring on new voices in the community. This year's survey is open now — help us do it again: bit.ly/4bXFBGH
1
4
3,175
We want to hear from you. Our Annual Reader Survey is officially live. It takes less than 5 minutes, and your answers directly shape the topics we cover, the writers we feature, and where TDS goes next. Take the survey: bit.ly/4bXFBGH
7
3,136
We've officially launched our 2026 Annual Reader Survey! We want to know how you use Towards Data Science, what you value most, and where we can improve. Help shape the future of TDS: bit.ly/4bXFBGH
2,594
Will you be at the 3rd International Conference on AI Sensors and Transducers next week in South Korea?
sciforum.net/event/AIS2026?s…
1
2,823
The 3rd International Conference on AI Sensors and Transducers will take place on 2–7 August 2026 in the beautiful city of Jeju, Korea. Building on the achievements of the 1st edition (Singapore, 2024) and the 2nd edition (Malaysia, 2025), this event will continue to set a high standard in the field. sciforum.net/event/AIS2026?s…
1
3
2,155
The 3rd international conference on AI sensors and transducers is set take place August 2-7 in Jeju, South Korea. Find out more here: sciforum.net/event/AIS2026?s…
1
1,932
Context switching can reduce efficiency by up to 40%. Now imagine applying that tax to every AI-assisted task in your workflow.
That's the hidden cost of the multi-tool approach most of us defaulted into. Not because we chose it — because the ecosystem grew faster than our workflows could adapt.
Worth a read if you're starting to feel like you spend more time managing AI than using it.
towardsdatascience.com/how-t…
4
2,217
RAG gives your LLM a snapshot. Live web search gives it a window.
The difference matters in production, especially when users are asking about things your vector store hasn't seen yet. Three integration patterns, real code, clear tradeoffs.
towardsdatascience.com/groun…
1
4
1,099
You're not spending too much time on AI. You're spending too much time managing AI tools.
There's a difference. The first is a capability problem. The second is a workflow problem — and it's the one most practitioners actually have.
Switching between ChatGPT, Claude, and five other interfaces doesn't make you more capable. It makes you a prompt logistics coordinator. This piece makes the case for why unified AI platforms are the next logical step: towardsdatascience.com/how-t…
1
4
1,345
Join the conversation with the leaders shaping the future of enterprise data in Atlanta. On Oct 1, COLLIDE 2026 will bring together Fortune 500 data and AI executives for a day of high-level networking and strategic insights.
datasciconnect.com/events/co…
2
1,067
Your LLM isn't hallucinating because it's broken. It's hallucinating because it's working with stale data. Here's how to fix it: towardsdatascience.com/groun…
2
9
1,323
Connect with the architects of the modern AI enterprise in the heart of New York City. On Sep 2, Data Science Connect's conference spotlights how industry leaders are building scalable, compliant, and intelligent systems that go beyond the buzz.
Register ➡️ datasciconnect.com/events/ai…
2
1,040
It's almost time for our first webinar! Join Dr. Kim Fessel as she welcomes Jess Ramos of Big Data Energy and Manish Patel of CData to dive into pipeline modernization in practice.
Don't miss it: streamyard.com/watch/yb5tV3S…
1
899
Most data teams know their pipelines need to evolve. The challenge is figuring out where to start, what to prioritize, and how to modernize without turning it into a six-month project. Join Dr. Kim Fessel as she welcomes Jess Ramos of Big Data Energy and Manish Patel of CData to dive into pipeline modernization in practice.
Save your seat for this live webinar: streamyard.com/watch/yb5tV3S…
1
1
1
1,060
In our upcoming live webinar, Dr. Kim Fessel will be joined by Jess Ramos of Big Data Energy and Manish Patel of CData to talk through what pipeline modernization actually looks like in practice. Discover how to assess which pipelines to modernize first - without a 6 month planning cycle.
Register for the webinar: streamyard.com/watch/yb5tV3S…
1
3
1,496
We're hosting our first live webinar on April 21! Jess Ramos of Big Data Energy and Manish Patel of CData will join our host, Dr. Kim Fessel, to discuss what it really takes to modernize your data pipelines.
Register to join us: streamyard.com/watch/yb5tV3S…
1
3
1,148
Get ready for COLLIDE 2026 on Oct 1 for a day of thought leadership and strategic insights focused on the "how" behind enterprise AI. This event brings together the architects of the modern AI enterprise to share their blueprints for success.
datasciconnect.com/events/co…
3
1,892
Move beyond the hype and see how Fortune 500 leaders are operationalizing AI at scale. Join our friends Data Science Connect at 'The AI Enterprise Conference' in NYC on Sep 2 to explore real cases, measurable ROI, and strategies leading orgs use.
Register 👉 datasciconnect.com/events/ai…
1
786
Take your locally parallelized Python code and run it on a cloud cluster with almost no changes. @taupirho explains how to adapt a script for distributed computing using the Ray framework.
towardsdatascience.com/ray-d…
3
1,122
Improve your understanding of retrieval-augmented generation for source code. Kenneth Leung uses @cursor_ai as a case study to show how a full pipeline is built for contextual awareness.
towardsdatascience.com/how-c…
8
1,313
Build a music recommendation system using audio embeddings. Luke Stuckey covers the process from data preparation and model training with CNNs to final application deployment.
towardsdatascience.com/how-c…
1
3
1,002
Causal inference can seem intimidating. Ross Lauterbach shares a clear starting point, walking through key methods without overly complex math or libraries, perfect for those new to the field.
towardsdatascience.com/causa…
9
1,242
Does access to a massive foundation model eliminate the need for in-house training, especially when data is scarce? Pushpak Bhoge puts this to the test.
towardsdatascience.com/sam-3…
1
8
2,353
Build a more complete statistical toolkit with this guide from @polaris000_soc. Get a clear breakdown of when to use different detection tools and understand the WHY behind each tool to apply them more effectively.
towardsdatascience.com/a-cas…
8
908
Stop writing blogs, and start building impact. We help you refine your technical work with editorial support and high-visibility tools. TDS is the home for authors who want to truly unpack complex topics.
Learn more 👉 bit.ly/TDSContributor
770
Understand the end-to-end RAG pipeline that enables coding agents like @cursor_ai to provide context-aware suggestions. Kenneth Leung shares a detailed walkthrough of how a codebase is indexed.
towardsdatascience.com/how-c…
1
12
1,278
Learn how to deploy and run CPU-intensive Python code on cloud-based clusters. @taupirho walks through using the Ray library to scale your applications from a local machine to a multi-server environment.
towardsdatascience.com/ray-d…
4
979
Learn how convolutional neural networks can understand musical similarity. Luke Stuckey explains how to create audio embeddings using contrastive learning and deploy them in a music recommendation app. 🎶
towardsdatascience.com/how-c…
1
7
898
Move beyond predictive modeling and learn to estimate the true impact of your decisions. Ross Lauterbach shares this accessible intro to Causal ML.
towardsdatascience.com/causa…
5
1,032
Move beyond the hype of general-purpose AI. Pushpak Bhoge investigates if a giant like SAM3 can outperform a small, specialized model with a 6-hour compute budget in an autonomous setting.
towardsdatascience.com/sam-3…
2
9
1,557
Day 1: Code. Day 2: Publish. 🚀 Level up your career by becoming a published TDS author. From senior engineers to researchers, our contributors are the voices shaping the future of AI.
Write for TDS: bit.ly/TDSContributor
2
7
928
Learn why correctly estimating statistical "noise" is crucial for trend detection. @polaris000_soc demonstrates how the t-statistic uses residuals to avoid the pitfalls of a naive z-score.
towardsdatascience.com/a-cas…
1
7
1,089
Move beyond the AI hype and see how the Fortune 500 is operationalizing data at scale. On Oct 1, the COLLIDE 2026 conference arrives in Atlanta to spotlight real-world enterprise use cases and the strategies leading orgs use to drive measurable ROI.
datasciconnect.com/events/co…
1
964
Examine the technical architecture behind modern coding agents. In a new article, Kenneth Leung dissects the systems that power @cursor_ai's ability to understand your entire project repository.
towardsdatascience.com/how-c…
1
2
4
1,152
Human-written, human-led. 🧠 We don’t publish AI-generated text. We’re looking for original insights, unique breakthroughs, and the human voice behind the code. Share your story with the TDS community.
bit.ly/TDSContributor
1
866
Learn to scale your Python applications from a single machine to a multi-server cloud cluster. In part 2 of his series on the Ray library, @taupirho shares a step-by-step guide for deploying and running CPU-intensive code on @awscloud.
towardsdatascience.com/ray-d…
1
4
1,071
🎶 Learn how to build a music recommendation system using deep learning. Luke Stuckey covers the entire pipeline: from training a CNN with contrastive learning, to deploying the final model in a functional web application.
towardsdatascience.com/how-c…
8
1,105
Discover why specialized models maintain a 30x speed advantage and offer greater cost-efficiency over foundation models like SAM3 in production. Pushpak Bhoge breaks down the numbers.
towardsdatascience.com/sam-3…
1
1,359
Stop posting to an empty room. Share your work with our global audience and build your authority on the world's leading data science platform.
bit.ly/TDSContributor
1
1
864
Stop relying on manual scripts and fragile transfers. Join our sister site, @TheNewStack, on Jan 29 to learn how API-driven integration and advanced reporting are revolutionizing Managed File Transfer.
bit.ly/fortra-mft-webinar
1
2
1,296
Reduce the runtime of your Python jobs by distributing them across a cloud cluster. @taupirho demonstrates how moving a task to a 6-node EC2 cluster with Ray achieved a 3x performance improvement.
towardsdatascience.com/ray-d…
3
1,001
Understand how coding agents like @cursor_ai index your codebase for relevant suggestions. Kenneth Leung's new article explains the RAG pipeline, from semantic chunking and embedding generation to keeping the index synchronized.
towardsdatascience.com/how-c…
12
1,157
See how a deep learning model for music similarity can be turned into a real-world tool. Luke Stuckey's debut TDS article shares a look at deploying the trained model into a functional song recommender web app.
towardsdatascience.com/how-c…
4
1,031