LLM Apps, Crucial Data Skills, Multi-Agent AI Systems, and Other July Must-Reads | by TDS Editors

Feeling inspired to write your first TDS post? We’re always open to contributions from new authors.

If it’s already summer where you live, we hope you’re making the most of the warm weather and (hopefully? maybe?) more relaxed daily rhythms. Learning never stops, of course—at least not for data scientists—so if your idea of a good time includes diving into new challenges and exploring cutting-edge tools and workflows, you’re in for a treat.

Our July highlights, made up of the articles that created the biggest splash among our readers last month, cover a wide range of practical topics—and many of them are geared towards helping you raise your own bar and expand your skill set. Let’s dive in!

Monthly Highlights

Building LLM Apps: A Clear Step-By-Step Guide
Many ML practitioners have great ideas for AI-based products, yet, as Almog Baku points out, “there are no established best practices, and often, pioneers are left with no clear roadmap, needing to reinvent the wheel or getting stuck.” Fortunately, that’s no longer the case, now that Almog has put together a blueprint for navigating the complex landscape of LLM-native development.
Multi AI Agent Systems 101
Soon after LLMs went mainstream, product engineers started to discover all the various pain points and bottlenecks they create. Mariya Mansurova’s recent guide introduces one of the most promising strategies for addressing these challenges: multi-agent AI systems, where teams of agents, each with their own specialized “skill,” can collaborate with each other.
The 5 Data Science Skills You Can’t Ignore in 2024
In her excellent career-focused roundup, Sara Nóbrega observes that “while universities and formal education provide some essential skills, they often do not prepare students with the practical know-how needed in companies.” Sara aims to fill in this gap with recommendations for five areas data scientists should focus on in order to thrive in today’s job market.
17 (Advanced) RAG Techniques to Turn Your LLM App Prototype into a Production-Ready Solution
For a one-stop, comprehensive resource you can refer to whenever you need to tweak, refine, or upgrade your retrieval-augmented generation system, make sure to bookmark Dominik Polzer’s recent contribution, which goes well beyond the basics to cover metadata, query routing, sentence-window retrieval, and much more.
Fine-Tune Smaller Transformer Models: Text Classification
We round out our monthly lineup with a standout project walkthrough, courtesy of Ida Silfverskiöld: it patiently outlines the process of fine-tuning a smaller transformer model for an NLP task, working with a pre-trained encoder model with binary classes to identify clickbait vs. factual articles.